diff --git a/Readability.js b/Readability.js index f47fe36..9c0840c 100644 --- a/Readability.js +++ b/Readability.js @@ -716,8 +716,8 @@ Readability.prototype = { // Add a point for the paragraph itself as a base. contentScore += 1; - // Add points for any commas within this paragraph. - contentScore += innerText.split(',').length; + // Add points for any commas within this paragraph. Up to 60 points. + contentScore += Math.min(innerText.split(',').length, 60); // For every 100 characters in this paragraph, add another point. Up to 3 points. contentScore += Math.min(Math.floor(innerText.length / 100), 3); diff --git a/test/test-pages/clean-links/expected.html b/test/test-pages/clean-links/expected.html index 1acdbd2..94657d2 100644 --- a/test/test-pages/clean-links/expected.html +++ b/test/test-pages/clean-links/expected.html @@ -1,264 +1,262 @@
- -

Study Webtext

-

Study Webtext

+

"Bartleby the Scrivener: A Story of Wall-Street " (1853) 
Herman Melville

-

To the story text without notes

-

Prepared by Ann +

Prepared by Ann Woodlief, Virginia Commonwealth University

-

Click on text in red for hypertext notes and questions
I am a rather elderly man. The nature of my avocations for the last thirty years has brought me into more than ordinary contact with what would seem an interesting and somewhat singular set of men of whom as yet nothing that I know of has ever been written:-- I mean the law-copyists or scriveners. I have known very many of them, professionally and privately, and if I pleased, could relate divers histories, at which good-natured gentlemen might smile, and sentimental souls might weep. But I waive the biographies of all other scriveners for a few passages in the life of Bartleby, who was a scrivener the strangest I ever saw or heard of. While of other law-copyists I might write the complete life, of Bartleby nothing of that sort can be done. I believe that no materials exist for a full and satisfactory biography of this man. It is an irreparable loss to literature. Bartleby was one of those beings of whom nothing is ascertainable, except from the original sources, and in his case those are very small. What my own astonished eyes saw of Bartleby, that is all I know of him, except, indeed, one vague report which will appear in the sequel. -

Ere introducing the scrivener, as he first appeared to me, it is fit I make some mention of myself, my employees, my business, my chambers, and general surroundings; because some such description is indispensable to an adequate understanding of the chief character about to be presented.

-

Imprimis: I am a man who, from his youth upwards, has been filled with a profound conviction that the easiest way of life is the best.. Hence, though I belong to a profession proverbially energetic and nervous, even to turbulence, at times, yet nothing of that sort have I ever suffered to invade my peace. I am one of those unambitious lawyers who never addresses a jury, or in any way draws down public applause; but in the cool tranquillity of a snug retreat, do a snug business among rich men's bonds and mortgages and title-deeds. The late John Jacob Astor, a personage little given to poetic enthusiasm, had no hesitation in pronouncing my first grand point to be prudence; my next, method. I do not speak it in vanity, but simply record the fact, that I was not unemployed in my profession by the last John Jacob Astor; a name which, I admit, I love to repeat, for it hath a rounded and orbicular sound to it, and rings like unto bullion. I will freely add, that I was not insensible to the late John Jacob Astor's good opinion.

-

Some time prior to the period at which this little history begins, my avocations had been largely increased. The good old office, now extinct in the State of New York, of a Master in Chancery, had been conferred upon me. It was not a very arduous office, but very pleasantly remunerative. I seldom lose my temper; much more seldom indulge in dangerous indignation at wrongs and outrages; but I must be permitted to be rash here and declare, that I consider the sudden and violent abrogation of the office of Master of Chancery, by the new Constitution, as a----premature act; inasmuch as I had counted upon a life-lease of the profits, whereas I only received those of a few short years. But this is by the way.

-

My chambers were up stairs at No.--Wall-street. At one end they looked upon the white wall of the interior of a spacious sky-light shaft, penetrating the building from top to bottom. This view might have been considered rather tame than otherwise, deficient in what landscape painters call "life." But if so, the view from the other end of my chambers offered, at least, a contrast, if nothing more. In that direction my windows commanded an unobstructed view of a lofty brick wall,black by age and everlasting shade; which wall required no spy-glass to bring out its lurking beauties, but for the benefit of all near-sighted spectators, was pushed up to within ten feet of my window panes. Owing to the great height of the surrounding buildings, and my chambers being on the second floor, the interval between this wall and mine not a little resembled a huge square cistern.

-

At the period just preceding the advent of Bartleby, I had two persons as copyists in my employment, and a promising lad as an office-boy. First, Turkey; second, Nippers; third, Ginger Nut.These may seem names, the like of which are not usually found in the Directory. In truth they were nicknames, mutually conferred upon each other by my three clerks, and were deemed expressive of their respective persons or characters. Turkey was a short, pursy Englishman of about my own age, that is, somewhere not far from sixty. In the morning, one might say, his face was of a fine florid hue, but after twelve o'clock, meridian-- his dinner hour-- it blazed like a grate full of Christmas coals; and continued blazing--but, as it were, with a gradual wane--till 6 o'clock, P.M. or thereabouts, after which I saw no more of the proprietor of the face, which gaining its meridian with the sun, seemed to set with it, to rise, culminate, and decline the following day, with the like regularity and undiminished glory. There are many singular coincidences I have known in the course of my life, not the least among which was the fact that exactly when Turkey displayed his fullest beams from his red and radiant countenance, just then, too, at the critical moment, began the daily period when I considered his business capacities as seriously disturbed for the remainder of the twenty-four hours. Not that he was absolutely idle, or averse to business then; far from it. The difficulty was, he was apt to be altogether too energetic. There was a strange, inflamed, flurried, flighty recklessness of activity about him. He would be incautious in dipping his pen into his inkstand. All his blots upon my documents, were dropped there after twelve o'clock, meridian. Indeed, not only would he be reckless and sadly given to making blots in the afternoon, but some days he went further, and was rather noisy. At such times, too, his face flamed with augmented blazonry, as if cannel coal had been heaped on anthracite. He made an unpleasant racket with his chair; spilled his sand-box; in mending his pens, impatiently split them all to pieces, and threw them on the floor in a sudden passion; stood up and leaned over his table, boxing his papers about in a most indecorous manner, very sad to behold in an elderly manlike him. Nevertheless, as he was in many ways a most valuable person to me, and all the time before twelve o'clock, meridian, was the quickest, steadiest creature too, accomplishing a great deal of work in a style not easy to be matched--for these reasons, I was willingto overlook his eccentricities, though indeed, occasionally, I remonstrated with him. I did this very gently, however, because, though the civilest, nay, the blandest and most reverential of men in the morning, yet in the afternoon he was disposed, upon provocation, to be slightly rash with his tongue, in fact, insolent. Now, valuing his morning services as I did, and resolved not to lose them; yet, at the same time made uncomfortable by his inflamed ways after twelve o'clock; and being a man of peace, unwilling by my admonitions to call forth unseemingly retorts from him; I took upon me, one Saturday noon (he was always worse on Saturdays), to hint to him, very kindly, that perhaps now that he was growing old, it might be well to abridge his labors; in short, he need not come to my chambers after twelve o'clock, but, dinner over, had best go home to his lodgings and rest himself till tea-time. But no; he insisted upon his afternoon devotions. His countenance became intolerably fervid, as he oratorically assured me--gesticulating with a long ruler at the other end of the room--that if his services in the morning were useful, how indispensible, then, in the afternoon?

-

"With submission, sir," said Turkey on this occasion, "I consider myself your right-hand man. In the morning I but marshal and deploy my columns; but in the afternoon I put myself at their head, and gallantly charge the foe, thus!"--and he made a violent thrust with the ruler.

-

"But the blots, Turkey," intimated I.

-

"True,--but, with submission, sir, behold these hairs! I am getting old. Surely, sir, a blot or two of a warm afternoon is not the page--is honorable. With submission, sir, we both are getting old."

-

This appeal to my fellow-feeling was hardly to be resisted. At all events, I saw that go he would not. So I made up my mind to let him stay, resolving, nevertheless, to see to it, that during the afternoon he had to do with my less important papers.

-

Nippers, the second on my list, was a whiskered, sallow, and, upon the whole, rather piratical-looking young man of about five and twenty. I always deemed him the victim of two evil powers-- ambition and indigestion. The ambition was evinced by a certain impatience of the duties of a mere copyist, an unwarrantable usurpation of strictly profession affairs, such as the original drawing up of legal documents. The indigestion seemed betokened in an occasional nervous testiness and grinning irritability, causing the teeth to audibly grind together over mistakes committed in copying; unnecessary maledictions, hissed, rather than spoken, in the heat of business; and especially by a continual discontent with the height of the table where he worked. Though of a very ingenious mechanical turn, Nippers could never get this table to suit him. He put chips under it, blocks of various sorts, bits of pasteboard, and at last went so far as to attempt an exquisite adjustment by final pieces of folded blotting-paper. But no invention would answer. If, for the sake of easing his back, he brought the table lid at a sharp angle well up towards his chin, and wrote there like a man using the steep roof of a Dutch house for his desk:--then he declared that it stopped the circulation in his arms. If now he lowered the table to his waistbands, and stooped over it in writing, then there was a sore aching in his back. In short, the truth of the matter was, Nippers knew not what he wanted. Or, if he wanted anything, it was to be rid of a scrivener's table altogether. Among the manifestations of his diseased ambition was a fondness he had for receiving visits from certain ambiguous-looking fellows in seedy coats, whom he called his clients. Indeed I was aware that not only was he, at times, considerable of a ward-politician, but he occasionally did a little businessat the Justices' courts, and was not unknown on the steps of the Tombs. I have good reason to believe, however, that one individual who called upon him at my chambers, and who, with a grand air, he insisted was his client, was no other than a dun, and the alleged title-deed, a bill. But with all his failings, and the annoyances he caused me, Nippers, like his compatriot Turkey, was a very useful man to me; wrote a neat, swift hand; and, when he chose, was not deficient in a gentlemanly sort of deportment. Added to this, he always dressedin a gentlemanly sort of way; and so, incidentally, reflected credit upon my chambers. Whereas with respect to Turkey, I had much ado to keep him from being a reproach to me. His clothes were apt to look oily and smell of eating-houses. He wore his pantaloons very loose and baggy in summer. His coats were execrable; his hat not to be handled. But while the hat was a thing of indifference to me, inasmuch as his natural civility and deference, as a dependent Englishman, always led him to doff it the moment he entered the room, yet his coat was another matter. Concerning his coats, I reasoned with him; but with no effect. The truth was, I suppose, that a man with so small an income, could not afford to sport such a lustrous face and a lustrous coat at one and the same time. As Nippers once observed, Turkey's money went chiefly for red ink. One winter day I presented Turkey with a highly-respectable looking coat of my own, a padded gray coat, of a most comfortable warmth, and which buttoned straight up from the knee to the neck. I thought Turkey would appreciate the favor, and abate his rashness and obstreperousness of afternoons. But no. I verily believe that buttoning himself up in so downy and blanket-like a coat had a pernicious effect upon him; upon the same principle that too much oats are bad for horses. In fact, precisely as a rash, restive horse is said to feel his oats, so Turkey felt his coat. It made him insolent. He was a man whom prosperity harmed.

-

Though concerning the self-indulgent habits of Turkey I had my own private surmises, yet touching Nippers I was well persuaded that whatever might be his faults in other respects, he was, at least, a temperate young man. But indeed, nature herself seemed to have been his vintner, and at his birth charged him so thoroughly with an irritable, brandy-like disposition, that all subsequent potations were needless. When I consider how, amid the stillness of my chambers, Nippers would sometimes impatiently rise from his seat, and stooping over his table, spread his arms wide apart, seize the whole desk, and move it, and jerk it, with a grim, grinding motion on the floor, as if the table were a perverse voluntary agent, intent on thwarting and vexing him; I plainly perceive that for Nippers, brandy and water were altogether superfluous.

-

It was fortunate for me that, owing to its course--indigestion--the irritability and consequent nervousness of Nippers, were mainly observable in the morning, while in the afternoon he was comparatively mild. So that Turkey's paroxysms only coming on about twelve o'clock, I never had to do with their eccentricities at one time. Their fits relieved each other like guards. When Nippers' was on, Turkey's was off, and vice versa. This was a good natural arrangement under the circumstances.

-

Ginger Nut, the third on my list, was a lad some twelve years old. His father was a carman, ambitious of seeing his son on the bench instead of a cart, before he died. So he sent him to my office as a student at law, errand boy, and cleaner and sweeper, at the rate of one dollar a week. He had a little desk to himself, but he did not use it much. Upon inspection, the drawer exhibited a great array of the shells of various sorts of nuts. Indeed, to this quick-witted youth the whole noble science of the law was contained in a nut-shell. Not the least among the employments of Ginger Nut, as well as one which he discharged with the most alacrity, was his duty as cake and apple purveyor for Turkey and Nippers. Copying law papers being proverbially a dry, husky sort of business, my two scriveners were fain to moisten their mouths very often with Spitzenbergs to be had at the numerous stalls nigh the Custom House and Post Office. Also, they sent Ginger Nut very frequently for that peculiar cake--small, flat, round, and very spicy--after which he had been named by them. Of a cold morning when business was but dull, Turkey would gobble up scores of these cakes, as if they were mere wafers--indeed they sell them at the rate of six or eight for a penny--the scrape of his pen blending with the crunching of the crisp particles in his mouth. Of all the fiery afternoon blunders and flurried rashnesses of Turkey, was his once moistening a ginger-cake between his lips, and clapping it on to a mortgage for a seal. I came within an ace of dismissing him then. But he mollified me by making an oriental bow, and saying--"With submission, sir, it was generous of me to find you in stationery on my own account."

-

Now my original business--that of a conveyancer and title hunter, and drawer-up of recondite documents of all sorts--was considerably increased by receiving the master's office. There was now great work for scriveners. Not only must I push the clerks already with me, but I must have additional help. In answer to my advertisement, a motionless young man one morning, stood upon my office threshold, the door being open, for it was summer. I can see that figure now--pallidly neat, pitiably respectable, incurably forlorn! It was Bartleby.

-

After a few words touching his qualifications, I engaged him, glad to have among my corps of copyists a man of so singularly sedate an aspect, which I thought might operate beneficially upon the flighty temper of Turkey, and the fiery one of Nippers.

-

I should have stated before that ground glass folding-doors divided my premises into two parts, one of which was occupied by my scriveners, the other by myself. According to my humor I threw open these doors, or closed them. I resolved to assign Bartleby a corner by the folding-doors, but on my side of them, so as to have this quiet man within easy call, in case any trifling thing was to be done. I placed his desk close up to a small side window in that part of the room, a window which originally had afforded a lateral view of certain grimy back-yards and bricks, but which, owing to subsequent erections, commanded at present no view at all, though it gave some light. Within three feet of the panes was a wall, and the light came down from far above, between two lofty buildings, as from a very small opening in a dome. Still further to a satisfactory arrangement, I procured a high green folding screen, which might entirely isolate Bartleby from my sight, though not remove him from my voice. And thus, in a manner, privacy and society were conjoined.

-

At first Bartleby did an extraordinary quantity of writing. As if long famishingfor something to copy, he seemed to gorge himself on my documents. There was no pause for digestion. He ran a day and night line, copying by sun-light and by candle-light. I should have been quite delighted with his application, had be been cheerfully industrious. But he wrote on silently, palely, mechanically.

-

It is, of course, an indispensable part of a scrivener's business to verify the accuracy of his copy, word by word. Where there are two or more scriveners in an office, they assist each other in this examination, one reading from the copy, the other holding the original. It is a very dull, wearisome, and lethargic affair. I can readily imagine that to some sanguine temperaments it would be altogether intolerable. For example, I cannot credit that the mettlesome poet Byron would have contentedly sat down with Bartleby to examine a law document of, say five hundred pages, closely written in a crimpy hand.

-

Now and then, in the haste of business, it had been my habit to assist in comparing some brief document myself, calling Turkey or Nippers for this purpose. One object I had in placing Bartleby so handy to me behind the screen, was to avail myself of his services on such trivial occasions. It was on the third day, I think, of his being with me, and before any necessity had arisen for having his own writing examined, that, being much hurried to complete a small affair I had in hand, I abruptly called to Bartleby. In my haste and natural expectancy of instant compliance, I sat with my head bent over the original on my desk, and my right hand sideways, and somewhat nervously extended with the copy, so that immediately upon emerging from his retreat, Bartleby might snatch it and proceed to business without the least delay.

-

In this very attitude did I sit when I called to him, rapidly stating what it was I wanted him to do--namely, to examine a small paper with me. Imagine my surprise, nay, my consternation, when without moving from his privacy, Bartleby in a singularly mild, firm voice, replied,"I would prefer not to."

-

I sat awhile in perfect silence, rallying my stunned faculties. Immediately it occurred to me that my ears had deceived me, or Bartleby had entirely misunderstood my meaning. I repeated my request in the clearest tone I could assume. But in quite as clear a one came the previous reply, "I would prefer not to."

-

"Prefer not to," echoed I, rising in high excitement, and crossing the room with a stride, "What do you mean? Are you moon-struck? I want you to help me compare this sheet here--take it," and I thrust it towards him.

-

"I would prefer not to," said he.

-

I looked at him steadfastly. His face was leanly composed; his gray eye dimly calm. Not a wrinkle of agitation rippled him. Had there been the least uneasiness, anger, impatience or impertinence in his manner; in other words, had there been any thing ordinarily human about him, doubtless I should have violently dismissed him from the premises. But as it was, I should have as soon thought of turning my pale plaster-of-paris bust of Cicero out of doors. I stood gazing at him awhile, as he went on with his own writing, and then reseated myself at my desk. This is very strange, thought I. What had one best do? But my business hurried me. I concluded to forget the matter for the present, reserving it for my future leisure. So calling Nippers from the other room, the paper was speedily examined.

-

A few days after this, Bartleby concluded four lengthy documents, being quadruplicates of a week's testimony taken before me in my High Court of Chancery. It became necessary to examine them. It was an important suit, and great accuracy was imperative. Having all things arranged I called Turkey, Nippers and Ginger Nut from the next room, meaning to place the four copies in the hands of my four clerks, while I should read from the original. Accordingly Turkey, Nippers and Ginger Nut had taken their seats in a row, each with his document in hand, when I called to Bartleby to join this interesting group.

-

"Bartleby! quick, I am waiting."

-

I heard a low scrape of his chair legs on the unscraped floor, and soon he appeared standing at the entrance of his hermitage.

-

"What is wanted?" said he mildly.

-

"The copies, the copies," said I hurriedly. "We are going to examine them. There"--and I held towards him the fourth quadruplicate.

-

"I would prefer not to," he said, and gently disappeared behind the screen.

-

For a few moments I was turned into a pillar of salt, standing at the head of my seated column of clerks. Recovering myself, I advanced towards the screen, and demanded the reason for such extraordinary conduct.

-

"Why do you refuse?"

-

"I would prefer not to."

-

With any other man I should have flown outright into a dreadful passion, scorned all further words, and thrust him ignominiously from my presence. But there was something about Bartleby that not only strangely disarmed me, but in a wonderful manner touched and disconcerted me. I began to reason with him.

-

"These are your own copies we are about to examine. It is labor saving to you, because one examination will answer for your four papers. It is common usage. Every copyist is bound to help examine his copy. Is it not so? Will you not speak? Answer!"

-

"I prefer not to," he replied in a flute-like tone. It seemed to me that while I had been addressing him, he carefully revolved every statement that I made; fully comprehended the meaning; could not gainsay the irresistible conclusion; but, at the same time, some paramount consideration prevailed with him to reply as he did.

-

"You are decided, then, not to comply with my request--a request made according to common usage and common sense?"

-

He briefly gave me to understand that on that point my judgment was sound. Yes: his decision was irreversible.

-

It is not seldom the case that when a man is browbeaten in some unprecedented and violently unreasonable way, he begins to stagger in his own plainest faith. He begins, as it were, vaguely to surmise that, wonderful as it may be, all the justice and all the reason is on the other side. Accordingly, if any disinterested persons are present, he turns to them for some reinforcement for his own faltering mind.

-

"Turkey," said I, "what do you think of this? Am I not right?"

-

"With submission, sir," said Turkey, with his blandest tone, "I think that you are."

-

"Nippers," said I, "what do you think of it?"

-

"I think I should kick him out of the office."

-

(The reader of nice perceptions will here perceive that, it being morning, Turkey's answer is couched in polite and tranquil terms, but Nippers replies in ill-tempered ones. Or, to repeat a previous sentence, Nipper's ugly mood was on duty, and Turkey's off.)

-

"Ginger Nut," said I, willing to enlist the smallest suffrage in my behalf, "what do you think of it?"

-

"I think, sir, he's a little luny," replied Ginger Nut, with a grin.

-

"You hear what they say," said I, turning towards the screen, "come forth and do your duty."

-

But he vouchsafed no reply. I pondered a moment in sore perplexity. But once more business hurried me. I determined again to postpone the consideration of this dilemma to my future leisure. With a little trouble we made out to examine the papers without Bartleby, though at every page or two, Turkey deferentially dropped his opinion that this proceeding was quite out of the common; while Nippers, twitching in his chair with a dyspeptic nervousness, ground out between his set teeth occasional hissing maledictions against the stubborn oaf behind the screen. And for his (Nipper's) part, this was the first and the last time he would do another man's business without pay.

-

Meanwhile Bartleby sat in his hermitage, oblivious to every thing but his own peculiar business there.

-

Some days passed, the scrivener being employed upon another lengthy work. His late remarkable conduct led me to regard his way narrowly. I observed that he never went to dinner; indeed that he never went any where. As yet I had never of my personal knowledge known him to be outside of my office. He was a perpetual sentry in the corner. At about eleven o'clock though, in the morning, I noticed that Ginger Nut would advance toward the opening in Bartleby's screen, as if silently beckoned thither by a gesture invisible to me where I sat. That boy would then leave the office jingling a few pence, and reappear with a handful of ginger-nuts which he delivered in the hermitage, receiving two of the cakes for his trouble.

-

He lives, then, on ginger-nuts, thought I; never eats a dinner, properly speaking; he must be a vegetarian then, but no; he never eats even vegetables, he eats nothing but ginger-nuts. My mind then ran on in reveries concerning the probable effects upon the human constitution of living entirely on ginger-nuts. Ginger-nuts are so called because they contain ginger as one of their peculiar constituents, and the final flavoring one. Now what was ginger? A hot, spicy thing. Was Bartleby hot and spicy? Not at all. Ginger, then, had no effect upon Bartleby. Probably he preferred it should have none.

-

Nothing so aggravates an earnest person as a passive resistance. If the individual so resisted be of a not inhumane temper, and the resisting one perfectly harmless in his passivity; then, in the better moods of the former, he will endeavor charitably to construe to his imagination what proves impossible to be solved by his judgment. Even so, for the most part, I regarded Bartleby and his ways. Poor fellow! thought I, he means no mischief; it is plain he intends no insolence; his aspect sufficiently evinces that his eccentricities are involuntary. He is useful to me. I can get along with him. If I turn him away, the chances are he will fall in with some less indulgent employer, and then he will be rudely treated, and perhaps driven forth miserably to starve. Yes. Here I can cheaply purchase a delicious self-approval. To befriend Bartleby; to humor him in his strange willfulness, will cost me little or nothing, while I lay up in my soul what will eventually prove a sweet morsel for my conscience. But this mood was not invariable with me. The passiveness of Bartleby sometimes irritated me. I felt strangely goaded on to encounter him in new opposition, to elicit some angry spark from him answerable to my own. But indeed I might as well have essayed to strike fire with my knuckles against a bit of Windsor soap. But one afternoon the evil impulse in me mastered me, and the following little scene ensued:

-

"Bartleby," said I, "when those papers are all copied, I will compare them with you."

-

"I would prefer not to."

-

"How? Surely you do not mean to persist in that mulish vagary?"

-

No answer.

-

I threw open the folding-doors near by, and turning upon Turkey and Nippers, exclaimed in an excited manner--

-

"He says, a second time, he won't examine his papers. What do you think of it, Turkey?"

-

It was afternoon, be it remembered. Turkey sat glowing like a brass boiler, his bald head steaming, his hands reeling among his blotted papers.

-

"Think of it?" roared Turkey; "I think I'll just step behind his screen, and black his eyes for him!"

-

So saying, Turkey rose to his feet and threw his arms into a pugilistic position. He was hurrying away to make good his promise, when I detained him, alarmed at the effect of incautiously rousing Turkey's combativeness after dinner.

-

"Sit down, Turkey," said I, "and hear what Nippers has to say. What do you think of it, Nippers? Would I not be justified in immediately dismissing Bartleby?"

-

"Excuse me, that is for you to decide, sir. I think his conduct quite unusual, and indeed unjust, as regards Turkey and myself. But it may only be a passing whim."

-

"Ah," exclaimed I, "you have strangely changed your mind then--you speak very gently of him now."

-

"All beer," cried Turkey; "gentleness is effects of beer--Nippers and I dined together to-day. You see how gentle I am, sir. Shall I go and black his eyes?"

-

"You refer to Bartleby, I suppose. No, not to-day, Turkey," I replied; "pray, put up your fists."

-

I closed the doors, and again advanced towards Bartleby. I felt additional incentives tempting me to my fate. I burned to be rebelled against again. I remembered that Bartleby never left the office.

-

"Bartleby," said I, "Ginger Nut is away; just step round to the Post Office, won't you? (it was but a three minutes walk,) and see if there is any thing for me."

-

"I would prefer not to."

-

"You will not?"

-

"I prefer not."

-

I staggered to my desk, and sat there in a deep study. My blind inveteracy returned. Was there any other thing in which I could procure myself to be ignominiously repulsed by this lean, penniless with?--my hired clerk? What added thing is there, perfectly reasonable, that he will be sure to refuse to do?

-

"Bartleby!"

-

No answer.

-

"Bartleby," in a louder tone.

-

No answer.

-

"Bartleby," I roared.

-

Like a very ghost, agreeably to the laws of magical invocation, at the third summons, he appeared at the entrance of his hermitage.

-

"Go to the next room, and tell Nippers to come to me."

-

"I prefer not to," he respectfully and slowly said, and mildly disappeared.

-

"Very good, Bartleby," said I, in a quiet sort of serenely severe self-possessed tone, intimating the unalterable purpose of some terrible retribution very close at hand. At the moment I half intended something of the kind. But upon the whole, as it was drawing towards my dinner-hour, I thought it best to put on my hat and walk home for the day, suffering much from perplexity and distress of mind.

-

Shall I acknowledge it? The conclusion of this whole business was that it soon became a fixed fact of my chambers, that a pale young scrivener, by the name of Bartleby, had a desk there; that he copied for me at the usual rate of four cents a folio (one hundred words); but he was permanently exempt from examining the work done by him, that duty being transferred to Turkey and Nippers, one of compliment doubtless to their superior acuteness; moreover, said Bartleby was never on any account to be dispatched on the most trivial errand of any sort; and that even if entreated to take upon him such a matter, it was generally understood that he would prefer not to--in other words, that he would refuse point-blank.

-

32 As days passed on, I became considerably reconciled to Bartleby. His steadiness, his freedom from all dissipation, his incessant industry (except when he chose to throw himself into a standing revery behind his screen), his great stillness, his unalterableness of demeanor under all circumstances, made him a valuable acquisition. One prime thing was this,--he was always there;--first in the morning, continually through the day, and the last at night. I had a singular confidence in his honesty. I felt my most precious papers perfectly safe in his hands. Sometimes to be sure I could not, for the very soul of me, avoid falling into sudden spasmodic passions with him. For it was exceeding difficult to bear in mind all the time those strange peculiarities, privileges, and unheard of exemptions, forming the tacit stipulations on Bartleby's part under which he remained in my office. Now and then, in the eagerness of dispatching pressing business, I would inadvertently summon Bartleby, in a short, rapid tone, to put his finger, say, on the incipient tie of a bit of red tape with which I was about compressing some papers. Of course, from behind the screen the usual answer, "I prefer not to," was sure to come; and then, how could a human creature with the common infirmities of our nature, refrain from bitterly exclaiming upon such perverseness--such unreasonableness. However, every added repulse of this sort which I received only tended to lessen the probability of my repeating the inadvertence.

-

Here is must be said, that according to the custom of most legal gentlemen occupying chambers in densely-populated law buildings, there were several keys to my door. One was kept by a woman residing in the attic, which person weekly scrubbed and daily swept and dusted my apartments. Another was kept by Turkey for convenience sake. The third I sometimes carried in my own pocket. The fourth I knew not who had.

-

Now, one Sunday morning I happened to go to Trinity Church, to hear a celebrated preacher, and finding myself rather early on the ground, I thought I would walk round to my chambers for a while. Luckily I had my key with me; but upon applying it to the lock, I found it resisted by something inserted from the inside. Quite surprised, I called out; when to my consternation a key was turned from within; and thrusting his lean visage at me, and holding the door ajar, the apparition of Bartleby appeared, in his shirt sleeves, and otherwise in a strangely tattered dishabille, saying quietly that he was sorry, but he was deeply engaged just then, and--preferred not admitting me at present. In a brief word or two, he moreover added, that perhaps I had better walk round the block two or three times, and by that time he would probably have concluded his affairs. Now, the utterly unsurmised appearance of Bartleby, tenanting my law-chambers of a Sunday morning, with his cadaverously gentlemanly nonchalance, yet withal firm and self-possessed, had such a strange effect upon me, that incontinently I slunk away from my own door, and did as desired. But not without sundry twinges of impotent rebellion against the mild effrontery of this unaccountable scrivener. Indeed, it was his wonderful mildness chiefly, which not only disarmed me, but unmanned me, as it were. For I consider that one, for the time, is a sort of unmanned when he tranquilly permits his hired clerk to dictate to him, and order him away from his own premises. Furthermore, I was full of uneasiness as to what Bartleby could possibly be doing in my office in his shirt sleeves, and in an otherwise dismantled condition of a Sunday morning. Was any thing amiss going on? Nay, that was out of the question. It was not to be thought of for a moment that Bartleby was an immoral person. But what could he be doing there?--copying? Nay again, whatever might be his eccentricities, Bartleby was an eminently decorous person. He would be the last man to sit down to his desk in any state approaching to nudity. Besides, it was Sunday; and there was something about Bartleby that forbade the supposition that we would by any secular occupation violate the proprieties of the day.

-

Nevertheless, my mind was not pacified; and full of a restless curiosity, at last I returned to the door. Without hindrance I inserted my key, opened it, and entered. Bartleby was not to be seen. I looked round anxiously, peeped behind his screen; but it was very plain that he was gone. Upon more closely examining the place, I surmised that for an indefinite period Bartleby must have ate, dressed, and slept in my office, and that too without plate, mirror, or bed. The cushioned seat of a rickety old sofa in one corner bore t faint impress of a lean, reclining form. Rolled away under his desk, I found a blanket; under the empty grate, a blacking box and brush; on a chair, a tin basin, with soap and a ragged towel; in a newspaper a few crumbs of ginger-nuts and a morsel of cheese. Yet, thought I, it is evident enough that Bartleby has been making his home here, keeping bachelor's hallall by himself. Immediately then the thought came sweeping across me, What miserable friendlessness and loneliness are here revealed! His poverty is great; but his solitude, how horrible! Think of it. Of a Sunday, Wall-street is deserted as Petra; and every night of every day it is an emptiness. This building too, which of week-days hums with industry and life, at nightfall echoes with sheer vacancy, and all through Sunday is forlorn. And here Bartleby makes his home; sole spectator of a solitude which he has seen all populous--a sort of innocent and transformed Marius brooding among the ruins of Carthage!

-

For the first time in my life a feeling of overpowering stinging melancholy seized me. Before, I had never experienced aught but a not-unpleasing sadness. The bond of a common humanity now drew me irresistibly to gloom. A fraternal melancholy! For both I and Bartleby were sons of Adam. I remembered the bright silks and sparkling faces I had seen that day in gala trim, swan-like sailing down the Mississippi of Broadway; and I contrasted them with the pallid copyist, and thought to myself, Ah, happiness courts the light, so we deem the world is gay; but misery hides aloof, so we deem that misery there is none. These sad fancyings-- chimeras, doubtless, of a sick and silly brain--led on to other and more special thoughts, concerning the eccentricities of Bartleby. Presentiments of strange discoveries hovered round me. The scrivener's pale form appeared to me laid out, among uncaring strangers, in its shivering winding sheet.

-

Suddenly I was attracted by Bartleby's closed desk, the key in open sight left in the lock.

-

I mean no mischief, seek the gratification of no heartless curiosity, thought I; besides, the desk is mine, and its contents too, so I will make bold to look within. Every thing was methodically arranged, the papers smoothly placed. The pigeon holes were deep, and removing the files of documents, I groped into their recesses. Presently I felt something there, and dragged it out. It was an old bandanna handkerchief, heavy and knotted. I opened it, and saw it was a savings' bank.

-

I now recalled all the quiet mysteries which I had noted in the man. I remembered that he never spoke but to answer; that though at intervals he had considerable time to himself, yet I had never seen him reading--no, not even a newspaper; that for long periods he would stand looking out, at his pale window behind the screen, upon the dead brick wall; I was quite sure he never visited any refectory or eating house; while his pale face clearly indicated that he never drank beer like Turkey, or tea and coffee even, like other men; that he never went any where in particular that I could learn; never went out for a walk, unless indeed that was the case at present; that he had declined telling who he was, or whence he came, or whether he had any relatives in the world; that though so thin and pale, he never complained of ill health. And more than all, I remembered a certain unconscious air of pallid--how shall I call it?--of pallid haughtiness, say, or rather an austere reserve about him, which had positively awed me into my tame compliance with his eccentricities, when I had feared to ask him to do the slightest incidental thing for me, even though I might know, from his long-continued motionlessness, that behind his screen he must be standing in one of those dead-wall reveries of his.

-

Revolving all these things, and coupling them with the recently discovered fact that he made my office his constant abiding place and home, and not forgetful of his morbid moodiness; revolving all these things, a prudential feeling began to steal over me. My first emotions had been those of pure melancholy and sincerest pity; but just in proportion as the forlornness of Bartleby grew and grew to my imagination, did that same melancholy merge into fear, that pity into repulsion. So true it is, and so terrible too, that up to a certain point the thought or sight of misery enlists our best affections; but, in certain special cases, beyond that point it does not. They err who would assert that invariably this is owing to the inherent selfishness of the human heart. It rather proceeds from a certain hopelessness of remedying excessive and organic ill. To a sensitive being, pity is not seldom pain. And when at last it is perceived that such pity cannot lead to effectual succor, common sense bids the soul be rid of it. What I saw that morning persuaded me that the scrivener was the victim of innate and incurable disorder. I might give alms to his body; but his body did not pain him; it was his soul that suffered, and his soul I could not reach.

-

I did not accomplish the purpose of going to Trinity Church that morning. Somehow, the things I had seen disqualified me for the time from church-going. I walked homeward, thinking what I would do with Bartleby. Finally, I resolvedupon this;--I would put certain calm questions to him the next morning, touching his history, &c., and if he declined to answer then openly and reservedly (and I supposed he would prefer not), then to give him a twenty dollar bill over and above whatever I might owe him, and tell him his services were no longer required; but that if in any other way I could assist him, I would be happy to do so, especially if he desired to return to his native place, wherever that might be, I would willingly help to defray the expenses. Moreover, if after reaching home, he found himself at any time in want of aid, a letter from him would be sure of a reply.

-

The next morning came.

-

"Bartleby," said I, gently calling to him behind the screen.

-

No reply.

-

"Bartleby," said I, in a still gentler tone, "come here; I am not going to ask you to do any thing you would prefer not to do--I simply wish to speak to you."

-

Upon this he noiselessly slid into view.

-

"Will you tell me, Bartleby, where you were born?"

-

"I would prefer not to."

-

"Will you tell me anything about yourself?"

-

"I would prefer not to."

-

"But what reasonable objection can you have to speak to me? I feel friendly towards you."

-

He did not look at me while I spoke, but kept his glance fixed upon my bust of Cicero, which as I then sat, was directly behind me, some six inches above my head. "What is your answer, Bartleby?" said I, after waiting a considerable time for a reply, during which his countenance remained immovable, only there was the faintest conceivable tremor of the white attenuated mouth.

-

"At present I prefer to give no answer," he said, and retired into his hermitage.

-

It was rather weak in me I confess, but his manner on this occasion nettled me. Not only did there seem to lurk in it a certain disdain, but his perverseness seemed ungrateful, considering the undeniable good usage and indulgence he had received from me.

-

Again I sat ruminating what I should do.Mortified as I was at his behavior, and resolved as I had been to dismiss him when I entered my office, nevertheless I strangely felt something superstitious knocking at my heart, and forbidding me to carry out my purpose, and denouncing me for a villain if I dared to breathe one bitter word against this forlornest of mankind. At last, familiarly drawing my chair behind his screen, I sat down and said: "Bartleby, never mind then about revealing your history; but let me entreat you, as a friend, to comply as far as may be with the usages of this office. Say now you will help to examine papers tomorrow or next day: in short, say now that in a day or two you will begin to be a little reasonable:--say so, Bartleby."

-

"At present I would prefer not to be a little reasonable was his idly cadaverous reply.,"

-

Just then the folding-doors opened, and Nippers approached. He seemed suffering from an unusually bad night's rest, induced by severer indigestion than common. He overheard those final words of Bartleby.

-

"Prefer not, eh?" gritted Nippers--"I'd prefer him, if I were you, sir," addressing me--"I'd prefer him; I'd give him preferences, the stubborn mule! What is it, sir, pray, that he prefers not to do now?"

-

Bartleby moved not a limb.

-

"Mr. Nippers," said I, "I'd prefer that you would withdraw for the present."

-

Somehow, of late I had got into the way of involuntary using this word "prefer" upon all sorts of not exactly suitable occasions. And I trembled to think that my contact with the scrivener had already and seriously affected me in a mental way. And what further and deeper aberration might it not yet produce? This apprehension had not been without efficacy in determining me to summary means.

-

As Nippers, looking very sour and sulky, was departing, Turkey blandly and deferentially approached.

-

"With submission, sir," said he, "yesterday I was thinking about Bartleby here, and I think that if he would but prefer to take a quart of good ale every day, it would do much towards mending him, and enabling him to assist in examining his papers."

-

"So you have got the word too," said I, slightly excited.

-

"With submission, what word, sir," asked Turkey, respectfully crowding himself into the contracted space behind the screen, and by so doing, making me jostle the scrivener. "What word, sir?"

-

"I would prefer to be left alone here," said Bartleby, as if offended at being mobbed in his privacy.

-

"That's the word, Turkey," said I--"that's it."

-

"Oh, prefer oh yes--queer word. I never use it myself. But, sir as I was saying, if he would but prefer--"

-

"Turkey," interrupted I, "you will please withdraw."

-

"Oh, certainly, sir, if you prefer that I should."

-

As he opened the folding-door to retire, Nippers at his desk caught a glimpse of me, and asked whether I would prefer to have a certain paper copied on blue paper or white. He did not in the least roguishly accent the word prefer. It was plain that it involuntarily rolled from his tongue. I thought to myself, surely I must get rid of a demented man, who already has in some degree turned the tongues, if not the heads of myself and clerks. But I thought it prudent not to break the dismission at once.

-

The next day I noticed that Bartleby did nothing but stand at his window in his dead-wall revery. Upon asking him why he did not write, he said that he had decided upon doing no more writing.

-

"Why, how now? what next?" exclaimed I, "do no more writing?"

-

"No more."

-

"And what is the reason?"

-

"Do you not see the reason for yourself," he indifferently replied.

-

I looked steadfastly at him, and perceived that his eyes looked dull and glazed. Instantly it occurred to me, that his unexampled diligence in copying by his dim window for the first few weeks of his stay with me might have temporarily impaired his vision.

-

I was touched. I said something in condolence with him. I hinted that of course he did wisely in abstaining from writing for a while; and urged him to embrace that opportunity of taking wholesome exercise in the open air. This, however, he did not do. A few days after this, my other clerks being absent, and being in a great hurry to dispatch certain letters by the mail, I thought that, having nothing else earthly to do, Bartleby would surely be less inflexible than usual, and carry these letters to the post-office. But he blankly declined. So, much to my inconvenience, I went myself.

-

Still added days went by. Whether Bartleby's eyes improved or not, I could not say. To all appearance, I thought they did. But when I asked him if they did, he vouchsafed no answer. At all events, he would do no copying. At last, in reply to my urgings, he informed me that he had permanently given up copying.

-

"What!" exclaimed I; "suppose your eyes should get entirely well- better than ever before--would you not copy then?"

-

"I have given up copying," he answered, and slid aside.

-

He remained as ever, a fixture in my chamber. Nay--if that were possible--he became still more of a fixture than before. What was to be done? He would do nothing in the office: why should he stay there? In plain fact, he had now become a millstone to me, not only useless as a necklace, but afflictive to bear. Yet I was sorry for him. I speak less than truth when I say that, on his own account, he occasioned me uneasiness. If he would but have named a single relative or friend, I would instantly have written, and urged their taking the poor fellow away to some convenient retreat. But he seemed alone, absolutely alone in the universe. A bit of wreck</font> in the mid Atlantic. At length, necessities connected with my business tyrannized over all other considerations. Decently as I could, I told Bartleby that in six days' time he must unconditionally leave the office. I warned him to take measures, in the interval, for procuring some other abode. I offered to assist him in this endeavor, if he himself would but take the first step towards a removal. "And when you finally quit me, Bartleby," added I, "I shall see that you go not away entirely unprovided. Six days from this hour, remember."

-

At the expiration of that period, I peeped behind the screen, and lo! Bartleby was there.

-

I buttoned up my coat, balanced myself; advanced slowly towards him, touched his shoulder, and said, "The time has come; you must quit this place; I am sorry for you; here is money; but you must go."

-

"I would prefer not," he replied, with his back still towards me.

-

"You must."

-

He remained silent.

-

Now I had an unbounded confidence in this man's common honesty. He had frequently restored to me six pences and shillings carelessly dropped upon the floor, for I am apt to be very reckless in such shirt-button affairs. The proceeding then which followed will not be deemed extraordinary. "Bartleby," said I, "I owe you twelve dollars on account; here are thirty-two; the odd twenty are yours.--Will you take it? and I handed the bills towards him.

-

But he made no motion.

-

"I will leave them here then," putting them under a weight on the table. Then taking my hat and cane and going to the door I tranquilly turned and added--"After you have removed your things from these offices, Bartleby, you will of course lock the door--since every one is now gone for the day but you--and if you please, slip your key underneath the mat, so that I may have it in the morning. I shall not see you again; so good-bye to you. If hereafter in your new place of abode I can be of any service to you, do not fail to advise me by letter. Good-bye, Bartleby, and fare you well."

-

But he answered not a word; like the last column of some ruined temple, he remained standing mute and solitary in the middle of the otherwise deserted room.

-

As I walked home in a pensive mood, my vanity got the better of my pity. I could not but highly plume myself on my masterly management in getting rid of Bartleby. Masterly I call it, and such it must appear to any dispassionate thinker. The beauty of my procedure seemed to consist in its perfect quietness. There was no vulgar bullying, no bravado of any sort, no choleric hectoring and striding to and fro across the apartment, jerking out vehement commands for Bartleby to bundle himself off with his beggarly traps. Nothing of the kind. Without loudly bidding Bartleby depart--as an inferior genius might have done--I assumed the ground that depart he must; and upon the assumption built all I had to say. The more I thought over my procedure, the more I was charmed with it. Nevertheless, next morning, upon awakening, I had my doubts,--I had somehow slept off the fumes of vanity. One of the coolest and wisest hours a man has, is just after he awakes in the morning. My procedure seemed as sagacious as ever,--but only in theory. How it would prove in practice--there was the rub. It was truly a beautiful thought to have assumed Bartleby's departure; but, after all, that assumption was simply my own, and none of Bartleby's. The great point was, not whether I had assumed that he would quit me, but whether he would prefer so to do. He was more a man of preferences than assumptions.

-

After breakfast, I walked down town, arguing the probabilities pro and con. One moment I thought it would prove a miserable failure, and Bartleby would be found all alive at my office as usual; the next moment it seemed certain that I should see his chair empty. And so I kept veering about. At the corner of Broadway and Canal- street, I saw quite an excited group of people standing in earnest conversation.

-

"I'll take odds he doesn't," said a voice as I passed.

-

"Doesn't go?--done!" said I, "put up your money."

-

I was instinctively putting my hand in my pocket to produce my own, when I remembered that this was an election day. The words I had overheard bore no reference to Bartleby, but to the success or non-success of some candidate for the mayoralty. In my intent frame of mind, I had, as it were, imagined that all Broadway shared in my excitement, and were debating the same question with me. I passed on, very thankful that the uproar of the street screened my momentary absent-mindedness.

-

As I had intended, I was earlier than usual at my office door. I stood listening for a moment. All was still. He must be gone. I tried the knob. The door was locked. Yes, my procedure had worked to a charm; he indeed must be vanished. Yet a certain melancholy mixed with this: I was almost sorry for my brilliant success. I was fumbling under the door mat for the key, which Bartleby was to have left there for me, when accidentally my knee knocked against a panel, producing a summoning sound, and in response a voice came to me from within--"Not yet; I am occupied."

-

It was Bartleby.

-

I was thunderstruck. For an instant I stood like the man who, pipe in mouth, was killed one cloudless afternoon long ago in Virginia, by summer lightning; at his own warm open window he was killed, and remained leaning out there upon the dreamy afternoon, till some one touched him, when he fell. "Not gone!" I murmured at last. But again obeying that wondrous ascendancy which the inscrutable scrivener had over me, and from which ascendancy, for all my chafing, I could not completely escape, I slowly went down stairs and out into the street, and while walking round the block, considered what I should next do in this unheard-of-perplexity. Turn the man out by an actual thrusting I could not; to drive him away by calling him hard names would not do; calling in the police was an unpleasant idea; and yet, permit him to enjoy his cadaverous triumph over me,--this too I could not think of. What was to be done? or, if nothing could be done, was there any thing further that I could assume in the matter? Yes, as before I had prospectively assumed that Bartleby would depart, so now I might retrospectively assume that departed he was. In the legitimate carrying out of this assumption, I might enter my office in a great hurry, and pretending not to see Bartleby at all, walk straight against him as if he were air. Such a proceeding would in a singular degree have the appearance of a home-thrust. It was hardly possible that Bartleby could withstand such an application of the doctrine of assumptions. But upon second thoughts the success of the plan seemed rather dubious. I resolved to argue the matter over with him again.

-

Bartleby," said I, entering the office, with a quietly severe expression. "I am seriously displeased. I am pained, Bartleby. I had thought better of you. I had imagined you of such a gentlemanly organization, that in any delicate dilemma a slight hint would suffice--in short, an assumption. But it appears I am deceived. Why," I added, unaffectedly starting, "you have not even touched the money yet," pointing to it, just where I had left it the evening previous.

-

He answered nothing.

-

"Will you, or will you not, quit me?" I now demanded in a sudden passion, advancing close to him.

-

"I would prefer not to quit you," he replied, gently emphasizing the not.

-

"What earthly right have you to stay here? do you pay any rent? Do you pay my taxes? Or is this property yours?"

-

He answered nothing.

-

"Are you ready to go on and write now? Are your eyes recovered? Could you copy a small paper for me this morning? or help examine a few lines? or step round to the post-office? In a word, will you do any thing at all, to give a coloring to your refusal to depart the premises?"

-

He silently retired into his hermitage.

-

I was now in such a state of nervous resentment that I thought it but prudentto check myself at present from further demonstrations. Bartleby and I were alone. I remembered the tragedy of the unfortunate Adams and the still more unfortunate Colt in the solitary office of the latter; and how poor Colt, being dreadfully incensed by Adams, and imprudently permitting himself to get wildly excited, was at unawares hurried into his fatal act--an act which certainly no man could possibly deplore more than the actor himself. Often it had occurred to me in my ponderings upon the subject, that had that altercation taken place in the public street, or at a private residence, it would not have terminated as it did. It was the circumstance of being alone in a solitary office, up stairs, of a building entirely unhallowed by humanizing domestic associations--an uncarpeted office, doubtless of a dusty, haggard sort of appearance;--this it must have been, which greatly helped to enhance the irritable desperation of the hapless Colt.

-

But when this old Adam of resentment rose in me and tempted me concerning Bartleby, I grappled him and threw him. How? Why, simply by recalling the divine injunction: "A new commandment give I unto you, that ye love one another." Yes, this it was that saved me. Aside from higher considerations, charity often operates as a vastly wise and prudent principle--a great safeguard to its possessor. Men have committed murder for jealousy's sake, and anger's sake, and hatred's sake, and selfishness' sake, and spiritual pride's sake; but no man that ever I heard of, ever committed a diabolical murder for sweet charity's sake. Mere self-interest, then, if no better motive can be enlisted, should, especially with high-tempered men, prompt all beings to charity and philanthropy. At any rate, upon the occasion in question, I strove to drown my exasperated feelings towards the scrivener by benevolently construing his conduct. Poor fellow, poor fellow! thought I, he don't mean any thing; and besides, he has seen hard times, and ought to be indulged.

-

I endeavored also immediately to occupy myself, and at the same time to comfort my despondency.I tried to fancy that in the course of the morning, at such time as might prove agreeable to him, Bartleby, of his own free accord, would emerge from his hermitage, and take up some decided line of march in the direction of the door. But no. Half-past twelve o'clock came; Turkey began to glow in the face, overturn his inkstand, and become generally obstreperous; Nippers abated down into quietude and courtesy; Ginger Nut munched his noon apple; and Bartleby remained standing at his window in one of his profoundest deadwall reveries. Will it be credited? Ought I to acknowledge it? That afternoon I left the office without saying one further word to him.

-

Some days now passed, during which, at leisure intervals I looked a little into Edwards on the Will," and "Priestly on Necessity." Under the circumstances, those books induced a salutary feeling. Gradually I slid into the persuasion that these troubles of mine touching the scrivener, had been all predestinated from eternity, and Bartleby was billeted upon me for some mysterious purpose of an all-wise Providence, which it was not for a mere mortal like me to fathom. Yes, Bartleby, stay there behind your screen, thought I; I shall persecute you no more; you are harmless and noiseless as any of these old chairs; in short, I never feel so private as when I know you are here. At least I see it, I feel it; I penetrate to the predestinated purpose of my life. I am content. Others may have loftier parts to enact; but my mission in this world, Bartleby, is to furnish you with office-room for such period as you may see fit to remain.

-

I believe that this wise and blessed frame of mind would have continued with me, had it not been for the unsolicited and uncharitable remarks obtruded upon me by my professional friends who visited the rooms. But thus it often is, that the constant friction of illiberal minds wears out at last the best resolves of the more generous. Though to be sure, when I reflected upon it, it was not strange that people entering my office should be struck by the peculiar aspect of the unaccountable Bartleby, and so be tempted to throw out some sinister observations concerning him. Sometimes an attorney having business with me, and calling at my office, and finding no one but the scrivener there, would undertake to obtain some sort of precise information from him touching my whereabouts; but without heeding his idle talk, Bartleby would remain standing immovable in the middle of the room. So after contemplating him in that position for a time, the attorney would depart, no wiser than he came.

-

Also, when a Reference was going on, and the room full of lawyers and witnesses and business was driving fast; some deeply occupied legal gentleman present, seeing Bartleby wholly unemployed, would request him to run round to his (the legal gentleman's) office and fetch some papers for him. Thereupon, Bartleby would tranquilly decline, and remain idle as before. Then the lawyer would give a great stare, and turn to me. And what could I say? At last I was made aware that all through the circle of my professional acquaintance, a whisper of wonder was running round, having reference to the strange creature I kept at my office. This worried me very much. And as the idea came upon me of his possibly turning out a long-lived man, and keep occupying my chambers, and denying my authority; and perplexing my visitors; and scandalizing my professional reputation; and casting a general gloom over the premises; keeping soul and body together to the last upon his savings (for doubtless he spent but half a dime a day), and in the end perhaps outlive me, and claim possession of my office by right of his perpetual occupancy: as all these dark anticipations crowded upon me more and more, and my friends continually intruded their relentless remarks upon the apparition in my room; a great change was wrought in me. I resolved to gather all my faculties together, and for ever rid me of this intolerable incubus.

-

Ere revolving any complicated project, however, adapted to this end, I first simply suggested to Bartleby the propriety of his permanent departure. In a calm and serious tone, I commended the idea to his careful and mature consideration. But having taken three days to meditate upon it, he apprised me that his original determination remained the same; in short, that he still preferred to abide with me.

-

What shall I do? I now said to myself, buttoning up my coat to the last button. What shall I do? what ought I to do? what does conscience say I should do with this man, or rather ghost. Rid myself of him, I must; go, he shall. But how? You will not thrust him, the poor, pale, passive mortal,--you will not thrust such a helpless creature out of your door? you will not dishonor yourself by such cruelty? No, I will not, I cannot do that. Rather would I let him live and die here, and then mason up his remains in the wall. What then will you do? For all your coaxing, he will not budge. Bribes he leaves under your own paperweight on your table; in short, it is quite plain that he prefers to cling to you.

-

Then something severe, something unusual must be done. What! surely you will not have him collared by a constable, and commit his innocent pallor to the common jail? And upon what ground could you procure such a thing to be done?--a vagrant, is he? What! he a vagrant, a wanderer, who refuses to budge? It is because he will not be a vagrant, then, that you seek to count him as a vagrant. That is too absurd. No visible means of support: there I have him. Wrong again: for indubitably he does support himself, and that is the only unanswerable proof that any man can show of his possessing the means so to do. No more then. Since he will not quit me, I must quit him. I will change my offices; I will move elsewhere; and give him fair notice, that if I find him on my new premises I will then proceed against him as a common trespasser.

-

Acting accordingly, next day I thus addressed him: "I find these chambers too far from the City Hall; the air is unwholesome. In a word, I propose to remove my offices next week, and shall no longer require your services. I tell you this now, in order that you may seek another place."

-

He made no reply, and nothing more was said.

-

On the appointed day I engaged carts and men, proceeded to my chambers, and having but little furniture, every thing was removed in a few hours. Throughout, the scrivener remained standing behind the screen, which I directed to be removed the last thing. It was withdrawn; and being folded up like a huge folio, left him the motionless occupant of a naked room. I stood in the entry watching him a moment, while something from within me upbraided me.

-

I re-entered, with my hand in my pocket--and--and my heart in my mouth.

-

"Good-bye, Bartleby; I am going--good-bye, and God some way bless you; and take that," slipping something in his hand. But it dropped to the floor, and then,--strange to say--I tore myself from him whom I had so longed to be rid of.

-

Established in my new quarters, for a day or two I kept the door locked, and started at every footfall in the passages. When I returned to my rooms after any little absence, I would pause at the threshold for an instant, and attentively listen, ere applying my key. But these fears were needless. Bartleby never came nigh me.

-

I thought all was going well, when a perturbed looking stranger visited me, inquiring whether I was the person who had recently occupied rooms at No.--Wall-street.

-

Full of forebodings, I replied that I was.

-

"Then, sir," said the stranger, who proved a lawyer, "you are responsible for the man you left there. He refuses to do any copying; he refuses to do any thing; he says he prefers not to; and he refuses to quit the premises."

-

"I am very sorry, sir," said I, with assumed tranquillity, but an inward tremor, "but, really, the man you allude to is nothing to me --he is no relation or apprentice of mine, that you should hold me responsible for him."

-

"In mercy's name, who is he?"

-

"I certainly cannot inform you. I know nothing about him. Formerly I employed him as a copyist; but he has done nothing for me now for some time past."

-

"I shall settle him then,--good morning, sir."

-

Several days passed, and I heard nothing more; and though I often felt a charitable prompting to call at the place and see poor Bartleby, yet a certain squeamishness of I know not what withheld me.

-

All is over with him, by this time, thought I at last, when through another week no further intelligence reached me. But coming to my room the day after, I found several persons waiting at my door in a high state of nervous excitement.

-

"That's the man--here he comes," cried the foremost one, whom recognized as the lawyer who had previously called upon me alone.

-

"You must take him away, sir, at once," cried a portly person among them, advancing upon me, and whom I knew to be the landlord of No.--Wall-street. "These gentlemen, my tenants, cannot stand it any longer; Mr. B--" pointing to the lawyer, "has turned him out of his room, and he now persists in haunting the buildinggenerally, sitting upon the banisters of the stairs by day, and sleeping in the entry by night. Every body is concerned; clients are leaving the offices; some fears are entertained of a mob; something you must do, and that without delay."

-

Aghast at this torment, I fell back before it, and would fain have locked myselfin my new quarters. In vain I persisted that Bartleby was nothing to me--no more than to any one else. In vain:--I was the last person known to have any thing to do with him, and they held me to the terrible account. Fearful then of being exposed in the papers (as one person present obscurely threatened) I considered the matter, and at length said, that if the lawyer would give me a confidential interview with the scrivener, in his (the lawyer's) own room, I would that afternoon strive my best to rid them of the nuisance they complained of.

-

Going up stairs to my old haunt, there was Bartleby silently sitting upon the banister at the landing.

-

"What are you doing here, Bartleby?" said I.

-

"Sitting upon the banister," he mildly replied.

-

I motioned him into the lawyer's room, who then left us.

-

"Bartleby," said I, "are you aware that you are the cause of great tribulation to me, by persisting in occupying the entry after being dismissed from the office?"

-

No answer.

-

"Now one of two things must take place. Either you must do something or something must be done to you. Now what sort of business would you like to engage in? Would you like to re-engage in copying for some one?"

-

"No; I would prefer not to make any change."

-

"Would you like a clerkship in a dry-goods store?"

-

"There is too much confinement about that. No, I would not like a clerkship; but I am not particular."

-

"Too much confinement," I cried, "why you keep yourself confined all the time!"

-

"I would prefer not to take a clerkship," he rejoined, as if to settle that little item at once.

-

"How would a bar-tender's business suit you? There is no trying of the eyesight in that."

-

"I would not like it at all; though, as I said before, I am not particular."

-

His unwonted wordiness inspirited me. I returned to the charge.

-

"Well then, would you like to travel through the country collecting bills for the merchants? That would improve your health."

-

"No, I would prefer to be doing something else."

-

"How then would going as a companion to Europe, to entertain some young gentleman with your conversation,--how would that suit you?"

-

"Not at all. It does not strike me that there is any thing definite about that. I like to be stationary. But I am not particular.

-

"Stationary you shall be then," I cried, now losing all patience, and for the first time in all my exasperating connection with him fairly flying into a passion. "If you do not go away from these premises before night, I shall feel bound--indeed I am bound--to-- to--to quit the premises myself!" I rather absurdly concluded, knowing not with what possible threat to try to frighten his immobility into compliance. Despairing of all further efforts, I was precipitately leaving him, when a final thought occurred to me--one which had not been wholly unindulged before.

-

"Bartleby," said I, in the kindest tone I could assume under such exciting circumstances, "will you go home with me now--not to my office, but my dwelling--and remain there till we can conclude upon some convenient arrangement for you at our leisure? Come, let us start now, right away."

-

"No: at present I would prefer not to make any change at all."

-

I answered nothing; but effectualy dodging every one by the suddenness and rapidity of my flight, rushed from the building, ran up Wall-street towards Broadway, and jumping into the first omnibus was soon removed from pursuit. As soon as tranquility returned I distinctly perceived that I had now done all that I possibly could, both in respect to the demands of the landlord and his tenants, and with regard to my own desire and sense of duty, to benefit Bartleby, and shield him from rude persecution. I now strove to be entirely care-free and quiescent; and my conscience justified me in the attempt; though indeed it was not so successful as I could have wished. So fearful was I of being again hunted out by the incensed landlord and his exasperated tenants, that, surrendering my business to Nippers, for a few days I drove about the upper part of the town and through the suburbs, in my rockaway; crossed over to Jersey City and Hoboken, and paid fugitive visits to Manhattanville and Astoria. In fact I almost lived in my rockaway for the time.

-

When again I entered my office, lo, a note from the landlord lay upon desk. opened it with trembling hands. informed me that writer had sent to police, and Bartleby removed the Tombs as a vagrant. Moreover, since I knew more about him than any one else, he wished me to appear at that place, and make a suitable statement of the facts. These tidings had a conflicting effect upon me. At first I was indignant; but at last almost approved. The landlord's energetic, summary disposition, had led him to adopt a procedure which I do not think I would have decided upon myself; and yet as a last resort, under such peculiar circumstances, it seemed the only plan.

-

As I afterwards learned, the poor scrivener, when told that he must be conducted to the Tombs, offered not the slightest obstacle, but in his pale unmoving way, silently acquiesced.

-

Some of the compassionate and curious bystanders joined the party; and headed by one of the constables arm in arm with Bartleby, the silent procession filed its way through all the noise, and heat, and joy of the roaring thoroughfares at noon.

-

The same day I received the note I went to the Tombs, or to speak more properly, the Halls of Justice. Seeking the right officer, I stated the purpose of my call, and was informed that the individual I described was indeed within. I then assured the functionary that Bartleby was a perfectly honest man, and greatly to be compassionated, however unaccountably eccentric. I narrated all I knew,and closed by suggesting the idea of letting him remain in as indulgent confinement as possible till something less harsh might be done--though indeed I hardly knew what. At all events, if nothing else could be decided upon, the alms-house must receive him. I then begged to have an interview.

-

Being under no disgraceful charge, and quite serene and harmless in all his ways, they had permitted him freely to wander about the prison, and especially in the inclosed grass-platted yards thereof. And so I found him there, standing all alone in the quietest of the yards, his face towards a high wall, while all around, from the narrow slits of the jail windows, I thought I saw peering out upon him the eyes of murderers and thieves.

-

"Bartleby!"

-

"I know you," he said, without looking round,--"and I want nothing to say to you."

-

"It was not I that brought you here, Bartleby," said I, keenly pained at his implied suspicion. "And to you, this should not be so vile a place. Nothing reproachful attaches to you by being here. And see, it is not so sad a place as one might think. Look, there is the sky, and here is the grass."

-

"I know where I am," he replied, but would say nothing more, and so I left him.

-

As I entered the corridor again, a broad meat-like man in an apron, accosted me, and jerking his thumb over his shoulder said--"Is that your friend?"

-

"Yes."

-

"Does he want to starve? If he does, let him live on the prison fare, that's all.

-

"Who are you?" asked I, not knowing what to make of such an unofficially speaking person in such a place.

-

"I am the grub-man. Such gentlemen as have friends here, hire me to provide them with something good to eat."

-

"Is this so?" said I, turning to the turnkey.

-

He said it was.

-

"Well then," said I, slipping some silver into the grub-man's hands (for so they called him). "I want you to give particular attention to my friend there; let him have the best dinner you can get. And you must be as polite to him as possible."

-

"Introduce me, will you?" said the grub-man, looking at me with an expression which seemed to say he was all impatience for an opportunity to give a specimen of his breeding.

-

Thinking it would prove of benefit to the scrivener, I acquiesced; and asking the grub-man his name, went up with him to Bartleby.

-

"Bartleby, this is a friend; you will find him very useful to you."

-

"Your sarvant, sir, your sarvant," said the grub-man, making a low salutation behind his apron. "Hope you find it pleasant here, sir;--spacious grounds--cool apartments, sir--hope you'll stay with us some time--try to make it agreeable. What will you have for dinner today?"

-

"I prefer not to dine to-day," said Bartleby, turning away. "It would disagree with me; I am unused to dinners." So saying he slowly moved to the other side of the inclosure, and took up a position fronting the dead-wall.

-

"How's this?" said the grub-man, addressing me with a stare of astonishment. "He's odd, aint he?"

-

"I think he is a little deranged," said I, sadly.

-

"Deranged? deranged is it? Well now, upon my word, I thought that friend of yourn was a gentleman forger; they are always pale and genteel-like, them forgers. I can't help pity 'em--can't help it, sir. Did you know Monroe Edwards?" he added touchingly, and paused. Then, laying his hand pityingly on my shoulder, sighed, "he died of consumption at Sing-Sing. so you weren't acquainted with Monroe?"

-

"No, I was never socially acquainted with any forgers. But I cannot stop longer. Look to my friend yonder. You will not lose by it. I will see you again."

-

Some few days after this, I again obtained admission to the Tombs, and went through the corridors in quest of Bartleby; but without finding him.

-

"I saw him coming from his cell not long ago," said a turnkey, "may be he's gone to loiter in the yards."

-

So I went in that direction.

-

"Are you looking for the silent man?" said another turnkey passing me. "Yonder he lies--sleeping in the yard there. 'Tis not twenty minutes since I saw him lie down."

-

The yard was entirely quiet. It was not accessible to the common prisoners. The surrounding walls, of amazing thickness, kept off all sound behind them. The Egyptian character of the masonry weighed upon me with its gloom. But a soft imprisoned turf grew under foot. The heart of the eternal pyramids, it seemed, wherein, by some strange magic, through the clefts, grass-seed, dropped by birds, had sprung.

-

Strangely huddled at the base of the wall, his knees drawn up, and lying on his side, his head touching the cold stones, I saw the wasted Bartleby. But nothing stirred. I paused; then went close up to him; stooped over, and saw that his dim eyes were open; otherwise he seemed profoundly sleeping. Something prompted me to touch him. I felt his hand, when a tingling shiver ran up my arm and down my spine to my feet.

-

The round face of the grub-man peered upon me now. "His dinner is ready. Won't he dine to-day, either? Or does he live without dining?"

-

"Lives without dining," said I, and closed the eyes.

-

"Eh!--He's asleep, aint he?"

-

"With kings and counsellors," murmured I.

-

* * * * * * * *

-

There would seem little need for proceeding further in this history. Imagination will readily supply the meagre recital of poor Bartleby's interment. But ere parting with the reader, let me say, that if this little narrative has sufficiently interested him, to awaken curiosity as to who Bartleby was, and what manner of life he led prior to the present narrator's making his acquaintance, I can only reply, that in such curiosity I fully share, but am wholly unable to gratify it. Yet here I hardly know whether I should divulge one little item of rumor, which came to my ear a few months after the scrivener's decease. Upon what basis it rested, I could never ascertain; and hence how true it is I cannot now tell. But inasmuch as this vague report has not been without a certain strange suggestive interest to me, however said, it may prove the same with some others; and so I will briefly mention it. The report was this: that Bartleby had been a subordinate clerk in the Dead Letter Office at Washington, from which he had been suddenly removed by a change in the administration. When I think over this rumor, I cannot adequately express the emotions which seize me. Dead letters! does it not sound like dead men? Conceive a man by nature and misfortune prone to a pallid hopelessness, can any business seem more fitted to heighten it than that of continually handling these dead letters and assorting them for the flames? For by the cart-load they are annually burned. Sometimes from out the folded paper the pale clerk takes a ring:--the bank-note sent in swiftest charity:--he whom it would relieve, nor eats nor hungers any more; pardon for those who died despairing; hope for those who died unhoping; good tidings for those who died stifled by unrelieved calamities. On errands of life, these letters speed to death.

-

Ah Bartleby! Ah humanity!

- +
Click on text in red for hypertext notes and questions
I am a rather elderly man. The nature of my avocations for the last thirty years has brought me into more than ordinary contact with what would seem an interesting and somewhat singular set of men of whom as yet nothing that I know of has ever been written:-- I mean the law-copyists or scriveners. I have known very many of them, professionally and privately, and if I pleased, could relate divers histories, at which good-natured gentlemen might smile, and sentimental souls might weep. But I waive the biographies of all other scriveners for a few passages in the life of Bartleby, who was a scrivener the strangest I ever saw or heard of. While of other law-copyists I might write the complete life, of Bartleby nothing of that sort can be done. I believe that no materials exist for a full and satisfactory biography of this man. It is an irreparable loss to literature. Bartleby was one of those beings of whom nothing is ascertainable, except from the original sources, and in his case those are very small. What my own astonished eyes saw of Bartleby, that is all I know of him, except, indeed, one vague report which will appear in the sequel. +

Ere introducing the scrivener, as he first appeared to me, it is fit I make some mention of myself, my employees, my business, my chambers, and general surroundings; because some such description is indispensable to an adequate understanding of the chief character about to be presented.

+

Imprimis: I am a man who, from his youth upwards, has been filled with a profound conviction that the easiest way of life is the best.. Hence, though I belong to a profession proverbially energetic and nervous, even to turbulence, at times, yet nothing of that sort have I ever suffered to invade my peace. I am one of those unambitious lawyers who never addresses a jury, or in any way draws down public applause; but in the cool tranquillity of a snug retreat, do a snug business among rich men's bonds and mortgages and title-deeds. The late John Jacob Astor, a personage little given to poetic enthusiasm, had no hesitation in pronouncing my first grand point to be prudence; my next, method. I do not speak it in vanity, but simply record the fact, that I was not unemployed in my profession by the last John Jacob Astor; a name which, I admit, I love to repeat, for it hath a rounded and orbicular sound to it, and rings like unto bullion. I will freely add, that I was not insensible to the late John Jacob Astor's good opinion.

+

Some time prior to the period at which this little history begins, my avocations had been largely increased. The good old office, now extinct in the State of New York, of a Master in Chancery, had been conferred upon me. It was not a very arduous office, but very pleasantly remunerative. I seldom lose my temper; much more seldom indulge in dangerous indignation at wrongs and outrages; but I must be permitted to be rash here and declare, that I consider the sudden and violent abrogation of the office of Master of Chancery, by the new Constitution, as a----premature act; inasmuch as I had counted upon a life-lease of the profits, whereas I only received those of a few short years. But this is by the way.

+

My chambers were up stairs at No.--Wall-street. At one end they looked upon the white wall of the interior of a spacious sky-light shaft, penetrating the building from top to bottom. This view might have been considered rather tame than otherwise, deficient in what landscape painters call "life." But if so, the view from the other end of my chambers offered, at least, a contrast, if nothing more. In that direction my windows commanded an unobstructed view of a lofty brick wall,black by age and everlasting shade; which wall required no spy-glass to bring out its lurking beauties, but for the benefit of all near-sighted spectators, was pushed up to within ten feet of my window panes. Owing to the great height of the surrounding buildings, and my chambers being on the second floor, the interval between this wall and mine not a little resembled a huge square cistern.

+

At the period just preceding the advent of Bartleby, I had two persons as copyists in my employment, and a promising lad as an office-boy. First, Turkey; second, Nippers; third, Ginger Nut.These may seem names, the like of which are not usually found in the Directory. In truth they were nicknames, mutually conferred upon each other by my three clerks, and were deemed expressive of their respective persons or characters. Turkey was a short, pursy Englishman of about my own age, that is, somewhere not far from sixty. In the morning, one might say, his face was of a fine florid hue, but after twelve o'clock, meridian-- his dinner hour-- it blazed like a grate full of Christmas coals; and continued blazing--but, as it were, with a gradual wane--till 6 o'clock, P.M. or thereabouts, after which I saw no more of the proprietor of the face, which gaining its meridian with the sun, seemed to set with it, to rise, culminate, and decline the following day, with the like regularity and undiminished glory. There are many singular coincidences I have known in the course of my life, not the least among which was the fact that exactly when Turkey displayed his fullest beams from his red and radiant countenance, just then, too, at the critical moment, began the daily period when I considered his business capacities as seriously disturbed for the remainder of the twenty-four hours. Not that he was absolutely idle, or averse to business then; far from it. The difficulty was, he was apt to be altogether too energetic. There was a strange, inflamed, flurried, flighty recklessness of activity about him. He would be incautious in dipping his pen into his inkstand. All his blots upon my documents, were dropped there after twelve o'clock, meridian. Indeed, not only would he be reckless and sadly given to making blots in the afternoon, but some days he went further, and was rather noisy. At such times, too, his face flamed with augmented blazonry, as if cannel coal had been heaped on anthracite. He made an unpleasant racket with his chair; spilled his sand-box; in mending his pens, impatiently split them all to pieces, and threw them on the floor in a sudden passion; stood up and leaned over his table, boxing his papers about in a most indecorous manner, very sad to behold in an elderly manlike him. Nevertheless, as he was in many ways a most valuable person to me, and all the time before twelve o'clock, meridian, was the quickest, steadiest creature too, accomplishing a great deal of work in a style not easy to be matched--for these reasons, I was willingto overlook his eccentricities, though indeed, occasionally, I remonstrated with him. I did this very gently, however, because, though the civilest, nay, the blandest and most reverential of men in the morning, yet in the afternoon he was disposed, upon provocation, to be slightly rash with his tongue, in fact, insolent. Now, valuing his morning services as I did, and resolved not to lose them; yet, at the same time made uncomfortable by his inflamed ways after twelve o'clock; and being a man of peace, unwilling by my admonitions to call forth unseemingly retorts from him; I took upon me, one Saturday noon (he was always worse on Saturdays), to hint to him, very kindly, that perhaps now that he was growing old, it might be well to abridge his labors; in short, he need not come to my chambers after twelve o'clock, but, dinner over, had best go home to his lodgings and rest himself till tea-time. But no; he insisted upon his afternoon devotions. His countenance became intolerably fervid, as he oratorically assured me--gesticulating with a long ruler at the other end of the room--that if his services in the morning were useful, how indispensible, then, in the afternoon?

+

"With submission, sir," said Turkey on this occasion, "I consider myself your right-hand man. In the morning I but marshal and deploy my columns; but in the afternoon I put myself at their head, and gallantly charge the foe, thus!"--and he made a violent thrust with the ruler.

+

"But the blots, Turkey," intimated I.

+

"True,--but, with submission, sir, behold these hairs! I am getting old. Surely, sir, a blot or two of a warm afternoon is not the page--is honorable. With submission, sir, we both are getting old."

+

This appeal to my fellow-feeling was hardly to be resisted. At all events, I saw that go he would not. So I made up my mind to let him stay, resolving, nevertheless, to see to it, that during the afternoon he had to do with my less important papers.

+

Nippers, the second on my list, was a whiskered, sallow, and, upon the whole, rather piratical-looking young man of about five and twenty. I always deemed him the victim of two evil powers-- ambition and indigestion. The ambition was evinced by a certain impatience of the duties of a mere copyist, an unwarrantable usurpation of strictly profession affairs, such as the original drawing up of legal documents. The indigestion seemed betokened in an occasional nervous testiness and grinning irritability, causing the teeth to audibly grind together over mistakes committed in copying; unnecessary maledictions, hissed, rather than spoken, in the heat of business; and especially by a continual discontent with the height of the table where he worked. Though of a very ingenious mechanical turn, Nippers could never get this table to suit him. He put chips under it, blocks of various sorts, bits of pasteboard, and at last went so far as to attempt an exquisite adjustment by final pieces of folded blotting-paper. But no invention would answer. If, for the sake of easing his back, he brought the table lid at a sharp angle well up towards his chin, and wrote there like a man using the steep roof of a Dutch house for his desk:--then he declared that it stopped the circulation in his arms. If now he lowered the table to his waistbands, and stooped over it in writing, then there was a sore aching in his back. In short, the truth of the matter was, Nippers knew not what he wanted. Or, if he wanted anything, it was to be rid of a scrivener's table altogether. Among the manifestations of his diseased ambition was a fondness he had for receiving visits from certain ambiguous-looking fellows in seedy coats, whom he called his clients. Indeed I was aware that not only was he, at times, considerable of a ward-politician, but he occasionally did a little businessat the Justices' courts, and was not unknown on the steps of the Tombs. I have good reason to believe, however, that one individual who called upon him at my chambers, and who, with a grand air, he insisted was his client, was no other than a dun, and the alleged title-deed, a bill. But with all his failings, and the annoyances he caused me, Nippers, like his compatriot Turkey, was a very useful man to me; wrote a neat, swift hand; and, when he chose, was not deficient in a gentlemanly sort of deportment. Added to this, he always dressedin a gentlemanly sort of way; and so, incidentally, reflected credit upon my chambers. Whereas with respect to Turkey, I had much ado to keep him from being a reproach to me. His clothes were apt to look oily and smell of eating-houses. He wore his pantaloons very loose and baggy in summer. His coats were execrable; his hat not to be handled. But while the hat was a thing of indifference to me, inasmuch as his natural civility and deference, as a dependent Englishman, always led him to doff it the moment he entered the room, yet his coat was another matter. Concerning his coats, I reasoned with him; but with no effect. The truth was, I suppose, that a man with so small an income, could not afford to sport such a lustrous face and a lustrous coat at one and the same time. As Nippers once observed, Turkey's money went chiefly for red ink. One winter day I presented Turkey with a highly-respectable looking coat of my own, a padded gray coat, of a most comfortable warmth, and which buttoned straight up from the knee to the neck. I thought Turkey would appreciate the favor, and abate his rashness and obstreperousness of afternoons. But no. I verily believe that buttoning himself up in so downy and blanket-like a coat had a pernicious effect upon him; upon the same principle that too much oats are bad for horses. In fact, precisely as a rash, restive horse is said to feel his oats, so Turkey felt his coat. It made him insolent. He was a man whom prosperity harmed.

+

Though concerning the self-indulgent habits of Turkey I had my own private surmises, yet touching Nippers I was well persuaded that whatever might be his faults in other respects, he was, at least, a temperate young man. But indeed, nature herself seemed to have been his vintner, and at his birth charged him so thoroughly with an irritable, brandy-like disposition, that all subsequent potations were needless. When I consider how, amid the stillness of my chambers, Nippers would sometimes impatiently rise from his seat, and stooping over his table, spread his arms wide apart, seize the whole desk, and move it, and jerk it, with a grim, grinding motion on the floor, as if the table were a perverse voluntary agent, intent on thwarting and vexing him; I plainly perceive that for Nippers, brandy and water were altogether superfluous.

+

It was fortunate for me that, owing to its course--indigestion--the irritability and consequent nervousness of Nippers, were mainly observable in the morning, while in the afternoon he was comparatively mild. So that Turkey's paroxysms only coming on about twelve o'clock, I never had to do with their eccentricities at one time. Their fits relieved each other like guards. When Nippers' was on, Turkey's was off, and vice versa. This was a good natural arrangement under the circumstances.

+

Ginger Nut, the third on my list, was a lad some twelve years old. His father was a carman, ambitious of seeing his son on the bench instead of a cart, before he died. So he sent him to my office as a student at law, errand boy, and cleaner and sweeper, at the rate of one dollar a week. He had a little desk to himself, but he did not use it much. Upon inspection, the drawer exhibited a great array of the shells of various sorts of nuts. Indeed, to this quick-witted youth the whole noble science of the law was contained in a nut-shell. Not the least among the employments of Ginger Nut, as well as one which he discharged with the most alacrity, was his duty as cake and apple purveyor for Turkey and Nippers. Copying law papers being proverbially a dry, husky sort of business, my two scriveners were fain to moisten their mouths very often with Spitzenbergs to be had at the numerous stalls nigh the Custom House and Post Office. Also, they sent Ginger Nut very frequently for that peculiar cake--small, flat, round, and very spicy--after which he had been named by them. Of a cold morning when business was but dull, Turkey would gobble up scores of these cakes, as if they were mere wafers--indeed they sell them at the rate of six or eight for a penny--the scrape of his pen blending with the crunching of the crisp particles in his mouth. Of all the fiery afternoon blunders and flurried rashnesses of Turkey, was his once moistening a ginger-cake between his lips, and clapping it on to a mortgage for a seal. I came within an ace of dismissing him then. But he mollified me by making an oriental bow, and saying--"With submission, sir, it was generous of me to find you in stationery on my own account."

+

Now my original business--that of a conveyancer and title hunter, and drawer-up of recondite documents of all sorts--was considerably increased by receiving the master's office. There was now great work for scriveners. Not only must I push the clerks already with me, but I must have additional help. In answer to my advertisement, a motionless young man one morning, stood upon my office threshold, the door being open, for it was summer. I can see that figure now--pallidly neat, pitiably respectable, incurably forlorn! It was Bartleby.

+

After a few words touching his qualifications, I engaged him, glad to have among my corps of copyists a man of so singularly sedate an aspect, which I thought might operate beneficially upon the flighty temper of Turkey, and the fiery one of Nippers.

+

I should have stated before that ground glass folding-doors divided my premises into two parts, one of which was occupied by my scriveners, the other by myself. According to my humor I threw open these doors, or closed them. I resolved to assign Bartleby a corner by the folding-doors, but on my side of them, so as to have this quiet man within easy call, in case any trifling thing was to be done. I placed his desk close up to a small side window in that part of the room, a window which originally had afforded a lateral view of certain grimy back-yards and bricks, but which, owing to subsequent erections, commanded at present no view at all, though it gave some light. Within three feet of the panes was a wall, and the light came down from far above, between two lofty buildings, as from a very small opening in a dome. Still further to a satisfactory arrangement, I procured a high green folding screen, which might entirely isolate Bartleby from my sight, though not remove him from my voice. And thus, in a manner, privacy and society were conjoined.

+

At first Bartleby did an extraordinary quantity of writing. As if long famishingfor something to copy, he seemed to gorge himself on my documents. There was no pause for digestion. He ran a day and night line, copying by sun-light and by candle-light. I should have been quite delighted with his application, had be been cheerfully industrious. But he wrote on silently, palely, mechanically.

+

It is, of course, an indispensable part of a scrivener's business to verify the accuracy of his copy, word by word. Where there are two or more scriveners in an office, they assist each other in this examination, one reading from the copy, the other holding the original. It is a very dull, wearisome, and lethargic affair. I can readily imagine that to some sanguine temperaments it would be altogether intolerable. For example, I cannot credit that the mettlesome poet Byron would have contentedly sat down with Bartleby to examine a law document of, say five hundred pages, closely written in a crimpy hand.

+

Now and then, in the haste of business, it had been my habit to assist in comparing some brief document myself, calling Turkey or Nippers for this purpose. One object I had in placing Bartleby so handy to me behind the screen, was to avail myself of his services on such trivial occasions. It was on the third day, I think, of his being with me, and before any necessity had arisen for having his own writing examined, that, being much hurried to complete a small affair I had in hand, I abruptly called to Bartleby. In my haste and natural expectancy of instant compliance, I sat with my head bent over the original on my desk, and my right hand sideways, and somewhat nervously extended with the copy, so that immediately upon emerging from his retreat, Bartleby might snatch it and proceed to business without the least delay.

+

In this very attitude did I sit when I called to him, rapidly stating what it was I wanted him to do--namely, to examine a small paper with me. Imagine my surprise, nay, my consternation, when without moving from his privacy, Bartleby in a singularly mild, firm voice, replied,"I would prefer not to."

+

I sat awhile in perfect silence, rallying my stunned faculties. Immediately it occurred to me that my ears had deceived me, or Bartleby had entirely misunderstood my meaning. I repeated my request in the clearest tone I could assume. But in quite as clear a one came the previous reply, "I would prefer not to."

+

"Prefer not to," echoed I, rising in high excitement, and crossing the room with a stride, "What do you mean? Are you moon-struck? I want you to help me compare this sheet here--take it," and I thrust it towards him.

+

"I would prefer not to," said he.

+

I looked at him steadfastly. His face was leanly composed; his gray eye dimly calm. Not a wrinkle of agitation rippled him. Had there been the least uneasiness, anger, impatience or impertinence in his manner; in other words, had there been any thing ordinarily human about him, doubtless I should have violently dismissed him from the premises. But as it was, I should have as soon thought of turning my pale plaster-of-paris bust of Cicero out of doors. I stood gazing at him awhile, as he went on with his own writing, and then reseated myself at my desk. This is very strange, thought I. What had one best do? But my business hurried me. I concluded to forget the matter for the present, reserving it for my future leisure. So calling Nippers from the other room, the paper was speedily examined.

+

A few days after this, Bartleby concluded four lengthy documents, being quadruplicates of a week's testimony taken before me in my High Court of Chancery. It became necessary to examine them. It was an important suit, and great accuracy was imperative. Having all things arranged I called Turkey, Nippers and Ginger Nut from the next room, meaning to place the four copies in the hands of my four clerks, while I should read from the original. Accordingly Turkey, Nippers and Ginger Nut had taken their seats in a row, each with his document in hand, when I called to Bartleby to join this interesting group.

+

"Bartleby! quick, I am waiting."

+

I heard a low scrape of his chair legs on the unscraped floor, and soon he appeared standing at the entrance of his hermitage.

+

"What is wanted?" said he mildly.

+

"The copies, the copies," said I hurriedly. "We are going to examine them. There"--and I held towards him the fourth quadruplicate.

+

"I would prefer not to," he said, and gently disappeared behind the screen.

+

For a few moments I was turned into a pillar of salt, standing at the head of my seated column of clerks. Recovering myself, I advanced towards the screen, and demanded the reason for such extraordinary conduct.

+

"Why do you refuse?"

+

"I would prefer not to."

+

With any other man I should have flown outright into a dreadful passion, scorned all further words, and thrust him ignominiously from my presence. But there was something about Bartleby that not only strangely disarmed me, but in a wonderful manner touched and disconcerted me. I began to reason with him.

+

"These are your own copies we are about to examine. It is labor saving to you, because one examination will answer for your four papers. It is common usage. Every copyist is bound to help examine his copy. Is it not so? Will you not speak? Answer!"

+

"I prefer not to," he replied in a flute-like tone. It seemed to me that while I had been addressing him, he carefully revolved every statement that I made; fully comprehended the meaning; could not gainsay the irresistible conclusion; but, at the same time, some paramount consideration prevailed with him to reply as he did.

+

"You are decided, then, not to comply with my request--a request made according to common usage and common sense?"

+

He briefly gave me to understand that on that point my judgment was sound. Yes: his decision was irreversible.

+

It is not seldom the case that when a man is browbeaten in some unprecedented and violently unreasonable way, he begins to stagger in his own plainest faith. He begins, as it were, vaguely to surmise that, wonderful as it may be, all the justice and all the reason is on the other side. Accordingly, if any disinterested persons are present, he turns to them for some reinforcement for his own faltering mind.

+

"Turkey," said I, "what do you think of this? Am I not right?"

+

"With submission, sir," said Turkey, with his blandest tone, "I think that you are."

+

"Nippers," said I, "what do you think of it?"

+

"I think I should kick him out of the office."

+

(The reader of nice perceptions will here perceive that, it being morning, Turkey's answer is couched in polite and tranquil terms, but Nippers replies in ill-tempered ones. Or, to repeat a previous sentence, Nipper's ugly mood was on duty, and Turkey's off.)

+

"Ginger Nut," said I, willing to enlist the smallest suffrage in my behalf, "what do you think of it?"

+

"I think, sir, he's a little luny," replied Ginger Nut, with a grin.

+

"You hear what they say," said I, turning towards the screen, "come forth and do your duty."

+

But he vouchsafed no reply. I pondered a moment in sore perplexity. But once more business hurried me. I determined again to postpone the consideration of this dilemma to my future leisure. With a little trouble we made out to examine the papers without Bartleby, though at every page or two, Turkey deferentially dropped his opinion that this proceeding was quite out of the common; while Nippers, twitching in his chair with a dyspeptic nervousness, ground out between his set teeth occasional hissing maledictions against the stubborn oaf behind the screen. And for his (Nipper's) part, this was the first and the last time he would do another man's business without pay.

+

Meanwhile Bartleby sat in his hermitage, oblivious to every thing but his own peculiar business there.

+

Some days passed, the scrivener being employed upon another lengthy work. His late remarkable conduct led me to regard his way narrowly. I observed that he never went to dinner; indeed that he never went any where. As yet I had never of my personal knowledge known him to be outside of my office. He was a perpetual sentry in the corner. At about eleven o'clock though, in the morning, I noticed that Ginger Nut would advance toward the opening in Bartleby's screen, as if silently beckoned thither by a gesture invisible to me where I sat. That boy would then leave the office jingling a few pence, and reappear with a handful of ginger-nuts which he delivered in the hermitage, receiving two of the cakes for his trouble.

+

He lives, then, on ginger-nuts, thought I; never eats a dinner, properly speaking; he must be a vegetarian then, but no; he never eats even vegetables, he eats nothing but ginger-nuts. My mind then ran on in reveries concerning the probable effects upon the human constitution of living entirely on ginger-nuts. Ginger-nuts are so called because they contain ginger as one of their peculiar constituents, and the final flavoring one. Now what was ginger? A hot, spicy thing. Was Bartleby hot and spicy? Not at all. Ginger, then, had no effect upon Bartleby. Probably he preferred it should have none.

+

Nothing so aggravates an earnest person as a passive resistance. If the individual so resisted be of a not inhumane temper, and the resisting one perfectly harmless in his passivity; then, in the better moods of the former, he will endeavor charitably to construe to his imagination what proves impossible to be solved by his judgment. Even so, for the most part, I regarded Bartleby and his ways. Poor fellow! thought I, he means no mischief; it is plain he intends no insolence; his aspect sufficiently evinces that his eccentricities are involuntary. He is useful to me. I can get along with him. If I turn him away, the chances are he will fall in with some less indulgent employer, and then he will be rudely treated, and perhaps driven forth miserably to starve. Yes. Here I can cheaply purchase a delicious self-approval. To befriend Bartleby; to humor him in his strange willfulness, will cost me little or nothing, while I lay up in my soul what will eventually prove a sweet morsel for my conscience. But this mood was not invariable with me. The passiveness of Bartleby sometimes irritated me. I felt strangely goaded on to encounter him in new opposition, to elicit some angry spark from him answerable to my own. But indeed I might as well have essayed to strike fire with my knuckles against a bit of Windsor soap. But one afternoon the evil impulse in me mastered me, and the following little scene ensued:

+

"Bartleby," said I, "when those papers are all copied, I will compare them with you."

+

"I would prefer not to."

+

"How? Surely you do not mean to persist in that mulish vagary?"

+

No answer.

+

I threw open the folding-doors near by, and turning upon Turkey and Nippers, exclaimed in an excited manner--

+

"He says, a second time, he won't examine his papers. What do you think of it, Turkey?"

+

It was afternoon, be it remembered. Turkey sat glowing like a brass boiler, his bald head steaming, his hands reeling among his blotted papers.

+

"Think of it?" roared Turkey; "I think I'll just step behind his screen, and black his eyes for him!"

+

So saying, Turkey rose to his feet and threw his arms into a pugilistic position. He was hurrying away to make good his promise, when I detained him, alarmed at the effect of incautiously rousing Turkey's combativeness after dinner.

+

"Sit down, Turkey," said I, "and hear what Nippers has to say. What do you think of it, Nippers? Would I not be justified in immediately dismissing Bartleby?"

+

"Excuse me, that is for you to decide, sir. I think his conduct quite unusual, and indeed unjust, as regards Turkey and myself. But it may only be a passing whim."

+

"Ah," exclaimed I, "you have strangely changed your mind then--you speak very gently of him now."

+

"All beer," cried Turkey; "gentleness is effects of beer--Nippers and I dined together to-day. You see how gentle I am, sir. Shall I go and black his eyes?"

+

"You refer to Bartleby, I suppose. No, not to-day, Turkey," I replied; "pray, put up your fists."

+

I closed the doors, and again advanced towards Bartleby. I felt additional incentives tempting me to my fate. I burned to be rebelled against again. I remembered that Bartleby never left the office.

+

"Bartleby," said I, "Ginger Nut is away; just step round to the Post Office, won't you? (it was but a three minutes walk,) and see if there is any thing for me."

+

"I would prefer not to."

+

"You will not?"

+

"I prefer not."

+

I staggered to my desk, and sat there in a deep study. My blind inveteracy returned. Was there any other thing in which I could procure myself to be ignominiously repulsed by this lean, penniless with?--my hired clerk? What added thing is there, perfectly reasonable, that he will be sure to refuse to do?

+

"Bartleby!"

+

No answer.

+

"Bartleby," in a louder tone.

+

No answer.

+

"Bartleby," I roared.

+

Like a very ghost, agreeably to the laws of magical invocation, at the third summons, he appeared at the entrance of his hermitage.

+

"Go to the next room, and tell Nippers to come to me."

+

"I prefer not to," he respectfully and slowly said, and mildly disappeared.

+

"Very good, Bartleby," said I, in a quiet sort of serenely severe self-possessed tone, intimating the unalterable purpose of some terrible retribution very close at hand. At the moment I half intended something of the kind. But upon the whole, as it was drawing towards my dinner-hour, I thought it best to put on my hat and walk home for the day, suffering much from perplexity and distress of mind.

+

Shall I acknowledge it? The conclusion of this whole business was that it soon became a fixed fact of my chambers, that a pale young scrivener, by the name of Bartleby, had a desk there; that he copied for me at the usual rate of four cents a folio (one hundred words); but he was permanently exempt from examining the work done by him, that duty being transferred to Turkey and Nippers, one of compliment doubtless to their superior acuteness; moreover, said Bartleby was never on any account to be dispatched on the most trivial errand of any sort; and that even if entreated to take upon him such a matter, it was generally understood that he would prefer not to--in other words, that he would refuse point-blank.

+

32 As days passed on, I became considerably reconciled to Bartleby. His steadiness, his freedom from all dissipation, his incessant industry (except when he chose to throw himself into a standing revery behind his screen), his great stillness, his unalterableness of demeanor under all circumstances, made him a valuable acquisition. One prime thing was this,--he was always there;--first in the morning, continually through the day, and the last at night. I had a singular confidence in his honesty. I felt my most precious papers perfectly safe in his hands. Sometimes to be sure I could not, for the very soul of me, avoid falling into sudden spasmodic passions with him. For it was exceeding difficult to bear in mind all the time those strange peculiarities, privileges, and unheard of exemptions, forming the tacit stipulations on Bartleby's part under which he remained in my office. Now and then, in the eagerness of dispatching pressing business, I would inadvertently summon Bartleby, in a short, rapid tone, to put his finger, say, on the incipient tie of a bit of red tape with which I was about compressing some papers. Of course, from behind the screen the usual answer, "I prefer not to," was sure to come; and then, how could a human creature with the common infirmities of our nature, refrain from bitterly exclaiming upon such perverseness--such unreasonableness. However, every added repulse of this sort which I received only tended to lessen the probability of my repeating the inadvertence.

+

Here is must be said, that according to the custom of most legal gentlemen occupying chambers in densely-populated law buildings, there were several keys to my door. One was kept by a woman residing in the attic, which person weekly scrubbed and daily swept and dusted my apartments. Another was kept by Turkey for convenience sake. The third I sometimes carried in my own pocket. The fourth I knew not who had.

+

Now, one Sunday morning I happened to go to Trinity Church, to hear a celebrated preacher, and finding myself rather early on the ground, I thought I would walk round to my chambers for a while. Luckily I had my key with me; but upon applying it to the lock, I found it resisted by something inserted from the inside. Quite surprised, I called out; when to my consternation a key was turned from within; and thrusting his lean visage at me, and holding the door ajar, the apparition of Bartleby appeared, in his shirt sleeves, and otherwise in a strangely tattered dishabille, saying quietly that he was sorry, but he was deeply engaged just then, and--preferred not admitting me at present. In a brief word or two, he moreover added, that perhaps I had better walk round the block two or three times, and by that time he would probably have concluded his affairs. Now, the utterly unsurmised appearance of Bartleby, tenanting my law-chambers of a Sunday morning, with his cadaverously gentlemanly nonchalance, yet withal firm and self-possessed, had such a strange effect upon me, that incontinently I slunk away from my own door, and did as desired. But not without sundry twinges of impotent rebellion against the mild effrontery of this unaccountable scrivener. Indeed, it was his wonderful mildness chiefly, which not only disarmed me, but unmanned me, as it were. For I consider that one, for the time, is a sort of unmanned when he tranquilly permits his hired clerk to dictate to him, and order him away from his own premises. Furthermore, I was full of uneasiness as to what Bartleby could possibly be doing in my office in his shirt sleeves, and in an otherwise dismantled condition of a Sunday morning. Was any thing amiss going on? Nay, that was out of the question. It was not to be thought of for a moment that Bartleby was an immoral person. But what could he be doing there?--copying? Nay again, whatever might be his eccentricities, Bartleby was an eminently decorous person. He would be the last man to sit down to his desk in any state approaching to nudity. Besides, it was Sunday; and there was something about Bartleby that forbade the supposition that we would by any secular occupation violate the proprieties of the day.

+

Nevertheless, my mind was not pacified; and full of a restless curiosity, at last I returned to the door. Without hindrance I inserted my key, opened it, and entered. Bartleby was not to be seen. I looked round anxiously, peeped behind his screen; but it was very plain that he was gone. Upon more closely examining the place, I surmised that for an indefinite period Bartleby must have ate, dressed, and slept in my office, and that too without plate, mirror, or bed. The cushioned seat of a rickety old sofa in one corner bore t faint impress of a lean, reclining form. Rolled away under his desk, I found a blanket; under the empty grate, a blacking box and brush; on a chair, a tin basin, with soap and a ragged towel; in a newspaper a few crumbs of ginger-nuts and a morsel of cheese. Yet, thought I, it is evident enough that Bartleby has been making his home here, keeping bachelor's hallall by himself. Immediately then the thought came sweeping across me, What miserable friendlessness and loneliness are here revealed! His poverty is great; but his solitude, how horrible! Think of it. Of a Sunday, Wall-street is deserted as Petra; and every night of every day it is an emptiness. This building too, which of week-days hums with industry and life, at nightfall echoes with sheer vacancy, and all through Sunday is forlorn. And here Bartleby makes his home; sole spectator of a solitude which he has seen all populous--a sort of innocent and transformed Marius brooding among the ruins of Carthage!

+

For the first time in my life a feeling of overpowering stinging melancholy seized me. Before, I had never experienced aught but a not-unpleasing sadness. The bond of a common humanity now drew me irresistibly to gloom. A fraternal melancholy! For both I and Bartleby were sons of Adam. I remembered the bright silks and sparkling faces I had seen that day in gala trim, swan-like sailing down the Mississippi of Broadway; and I contrasted them with the pallid copyist, and thought to myself, Ah, happiness courts the light, so we deem the world is gay; but misery hides aloof, so we deem that misery there is none. These sad fancyings-- chimeras, doubtless, of a sick and silly brain--led on to other and more special thoughts, concerning the eccentricities of Bartleby. Presentiments of strange discoveries hovered round me. The scrivener's pale form appeared to me laid out, among uncaring strangers, in its shivering winding sheet.

+

Suddenly I was attracted by Bartleby's closed desk, the key in open sight left in the lock.

+

I mean no mischief, seek the gratification of no heartless curiosity, thought I; besides, the desk is mine, and its contents too, so I will make bold to look within. Every thing was methodically arranged, the papers smoothly placed. The pigeon holes were deep, and removing the files of documents, I groped into their recesses. Presently I felt something there, and dragged it out. It was an old bandanna handkerchief, heavy and knotted. I opened it, and saw it was a savings' bank.

+

I now recalled all the quiet mysteries which I had noted in the man. I remembered that he never spoke but to answer; that though at intervals he had considerable time to himself, yet I had never seen him reading--no, not even a newspaper; that for long periods he would stand looking out, at his pale window behind the screen, upon the dead brick wall; I was quite sure he never visited any refectory or eating house; while his pale face clearly indicated that he never drank beer like Turkey, or tea and coffee even, like other men; that he never went any where in particular that I could learn; never went out for a walk, unless indeed that was the case at present; that he had declined telling who he was, or whence he came, or whether he had any relatives in the world; that though so thin and pale, he never complained of ill health. And more than all, I remembered a certain unconscious air of pallid--how shall I call it?--of pallid haughtiness, say, or rather an austere reserve about him, which had positively awed me into my tame compliance with his eccentricities, when I had feared to ask him to do the slightest incidental thing for me, even though I might know, from his long-continued motionlessness, that behind his screen he must be standing in one of those dead-wall reveries of his.

+

Revolving all these things, and coupling them with the recently discovered fact that he made my office his constant abiding place and home, and not forgetful of his morbid moodiness; revolving all these things, a prudential feeling began to steal over me. My first emotions had been those of pure melancholy and sincerest pity; but just in proportion as the forlornness of Bartleby grew and grew to my imagination, did that same melancholy merge into fear, that pity into repulsion. So true it is, and so terrible too, that up to a certain point the thought or sight of misery enlists our best affections; but, in certain special cases, beyond that point it does not. They err who would assert that invariably this is owing to the inherent selfishness of the human heart. It rather proceeds from a certain hopelessness of remedying excessive and organic ill. To a sensitive being, pity is not seldom pain. And when at last it is perceived that such pity cannot lead to effectual succor, common sense bids the soul be rid of it. What I saw that morning persuaded me that the scrivener was the victim of innate and incurable disorder. I might give alms to his body; but his body did not pain him; it was his soul that suffered, and his soul I could not reach.

+

I did not accomplish the purpose of going to Trinity Church that morning. Somehow, the things I had seen disqualified me for the time from church-going. I walked homeward, thinking what I would do with Bartleby. Finally, I resolvedupon this;--I would put certain calm questions to him the next morning, touching his history, &c., and if he declined to answer then openly and reservedly (and I supposed he would prefer not), then to give him a twenty dollar bill over and above whatever I might owe him, and tell him his services were no longer required; but that if in any other way I could assist him, I would be happy to do so, especially if he desired to return to his native place, wherever that might be, I would willingly help to defray the expenses. Moreover, if after reaching home, he found himself at any time in want of aid, a letter from him would be sure of a reply.

+

The next morning came.

+

"Bartleby," said I, gently calling to him behind the screen.

+

No reply.

+

"Bartleby," said I, in a still gentler tone, "come here; I am not going to ask you to do any thing you would prefer not to do--I simply wish to speak to you."

+

Upon this he noiselessly slid into view.

+

"Will you tell me, Bartleby, where you were born?"

+

"I would prefer not to."

+

"Will you tell me anything about yourself?"

+

"I would prefer not to."

+

"But what reasonable objection can you have to speak to me? I feel friendly towards you."

+

He did not look at me while I spoke, but kept his glance fixed upon my bust of Cicero, which as I then sat, was directly behind me, some six inches above my head. "What is your answer, Bartleby?" said I, after waiting a considerable time for a reply, during which his countenance remained immovable, only there was the faintest conceivable tremor of the white attenuated mouth.

+

"At present I prefer to give no answer," he said, and retired into his hermitage.

+

It was rather weak in me I confess, but his manner on this occasion nettled me. Not only did there seem to lurk in it a certain disdain, but his perverseness seemed ungrateful, considering the undeniable good usage and indulgence he had received from me.

+

Again I sat ruminating what I should do.Mortified as I was at his behavior, and resolved as I had been to dismiss him when I entered my office, nevertheless I strangely felt something superstitious knocking at my heart, and forbidding me to carry out my purpose, and denouncing me for a villain if I dared to breathe one bitter word against this forlornest of mankind. At last, familiarly drawing my chair behind his screen, I sat down and said: "Bartleby, never mind then about revealing your history; but let me entreat you, as a friend, to comply as far as may be with the usages of this office. Say now you will help to examine papers tomorrow or next day: in short, say now that in a day or two you will begin to be a little reasonable:--say so, Bartleby."

+

"At present I would prefer not to be a little reasonable was his idly cadaverous reply.,"

+

Just then the folding-doors opened, and Nippers approached. He seemed suffering from an unusually bad night's rest, induced by severer indigestion than common. He overheard those final words of Bartleby.

+

"Prefer not, eh?" gritted Nippers--"I'd prefer him, if I were you, sir," addressing me--"I'd prefer him; I'd give him preferences, the stubborn mule! What is it, sir, pray, that he prefers not to do now?"

+

Bartleby moved not a limb.

+

"Mr. Nippers," said I, "I'd prefer that you would withdraw for the present."

+

Somehow, of late I had got into the way of involuntary using this word "prefer" upon all sorts of not exactly suitable occasions. And I trembled to think that my contact with the scrivener had already and seriously affected me in a mental way. And what further and deeper aberration might it not yet produce? This apprehension had not been without efficacy in determining me to summary means.

+

As Nippers, looking very sour and sulky, was departing, Turkey blandly and deferentially approached.

+

"With submission, sir," said he, "yesterday I was thinking about Bartleby here, and I think that if he would but prefer to take a quart of good ale every day, it would do much towards mending him, and enabling him to assist in examining his papers."

+

"So you have got the word too," said I, slightly excited.

+

"With submission, what word, sir," asked Turkey, respectfully crowding himself into the contracted space behind the screen, and by so doing, making me jostle the scrivener. "What word, sir?"

+

"I would prefer to be left alone here," said Bartleby, as if offended at being mobbed in his privacy.

+

"That's the word, Turkey," said I--"that's it."

+

"Oh, prefer oh yes--queer word. I never use it myself. But, sir as I was saying, if he would but prefer--"

+

"Turkey," interrupted I, "you will please withdraw."

+

"Oh, certainly, sir, if you prefer that I should."

+

As he opened the folding-door to retire, Nippers at his desk caught a glimpse of me, and asked whether I would prefer to have a certain paper copied on blue paper or white. He did not in the least roguishly accent the word prefer. It was plain that it involuntarily rolled from his tongue. I thought to myself, surely I must get rid of a demented man, who already has in some degree turned the tongues, if not the heads of myself and clerks. But I thought it prudent not to break the dismission at once.

+

The next day I noticed that Bartleby did nothing but stand at his window in his dead-wall revery. Upon asking him why he did not write, he said that he had decided upon doing no more writing.

+

"Why, how now? what next?" exclaimed I, "do no more writing?"

+

"No more."

+

"And what is the reason?"

+

"Do you not see the reason for yourself," he indifferently replied.

+

I looked steadfastly at him, and perceived that his eyes looked dull and glazed. Instantly it occurred to me, that his unexampled diligence in copying by his dim window for the first few weeks of his stay with me might have temporarily impaired his vision.

+

I was touched. I said something in condolence with him. I hinted that of course he did wisely in abstaining from writing for a while; and urged him to embrace that opportunity of taking wholesome exercise in the open air. This, however, he did not do. A few days after this, my other clerks being absent, and being in a great hurry to dispatch certain letters by the mail, I thought that, having nothing else earthly to do, Bartleby would surely be less inflexible than usual, and carry these letters to the post-office. But he blankly declined. So, much to my inconvenience, I went myself.

+

Still added days went by. Whether Bartleby's eyes improved or not, I could not say. To all appearance, I thought they did. But when I asked him if they did, he vouchsafed no answer. At all events, he would do no copying. At last, in reply to my urgings, he informed me that he had permanently given up copying.

+

"What!" exclaimed I; "suppose your eyes should get entirely well- better than ever before--would you not copy then?"

+

"I have given up copying," he answered, and slid aside.

+

He remained as ever, a fixture in my chamber. Nay--if that were possible--he became still more of a fixture than before. What was to be done? He would do nothing in the office: why should he stay there? In plain fact, he had now become a millstone to me, not only useless as a necklace, but afflictive to bear. Yet I was sorry for him. I speak less than truth when I say that, on his own account, he occasioned me uneasiness. If he would but have named a single relative or friend, I would instantly have written, and urged their taking the poor fellow away to some convenient retreat. But he seemed alone, absolutely alone in the universe. A bit of wreck</font> in the mid Atlantic. At length, necessities connected with my business tyrannized over all other considerations. Decently as I could, I told Bartleby that in six days' time he must unconditionally leave the office. I warned him to take measures, in the interval, for procuring some other abode. I offered to assist him in this endeavor, if he himself would but take the first step towards a removal. "And when you finally quit me, Bartleby," added I, "I shall see that you go not away entirely unprovided. Six days from this hour, remember."

+

At the expiration of that period, I peeped behind the screen, and lo! Bartleby was there.

+

I buttoned up my coat, balanced myself; advanced slowly towards him, touched his shoulder, and said, "The time has come; you must quit this place; I am sorry for you; here is money; but you must go."

+

"I would prefer not," he replied, with his back still towards me.

+

"You must."

+

He remained silent.

+

Now I had an unbounded confidence in this man's common honesty. He had frequently restored to me six pences and shillings carelessly dropped upon the floor, for I am apt to be very reckless in such shirt-button affairs. The proceeding then which followed will not be deemed extraordinary. "Bartleby," said I, "I owe you twelve dollars on account; here are thirty-two; the odd twenty are yours.--Will you take it? and I handed the bills towards him.

+

But he made no motion.

+

"I will leave them here then," putting them under a weight on the table. Then taking my hat and cane and going to the door I tranquilly turned and added--"After you have removed your things from these offices, Bartleby, you will of course lock the door--since every one is now gone for the day but you--and if you please, slip your key underneath the mat, so that I may have it in the morning. I shall not see you again; so good-bye to you. If hereafter in your new place of abode I can be of any service to you, do not fail to advise me by letter. Good-bye, Bartleby, and fare you well."

+

But he answered not a word; like the last column of some ruined temple, he remained standing mute and solitary in the middle of the otherwise deserted room.

+

As I walked home in a pensive mood, my vanity got the better of my pity. I could not but highly plume myself on my masterly management in getting rid of Bartleby. Masterly I call it, and such it must appear to any dispassionate thinker. The beauty of my procedure seemed to consist in its perfect quietness. There was no vulgar bullying, no bravado of any sort, no choleric hectoring and striding to and fro across the apartment, jerking out vehement commands for Bartleby to bundle himself off with his beggarly traps. Nothing of the kind. Without loudly bidding Bartleby depart--as an inferior genius might have done--I assumed the ground that depart he must; and upon the assumption built all I had to say. The more I thought over my procedure, the more I was charmed with it. Nevertheless, next morning, upon awakening, I had my doubts,--I had somehow slept off the fumes of vanity. One of the coolest and wisest hours a man has, is just after he awakes in the morning. My procedure seemed as sagacious as ever,--but only in theory. How it would prove in practice--there was the rub. It was truly a beautiful thought to have assumed Bartleby's departure; but, after all, that assumption was simply my own, and none of Bartleby's. The great point was, not whether I had assumed that he would quit me, but whether he would prefer so to do. He was more a man of preferences than assumptions.

+

After breakfast, I walked down town, arguing the probabilities pro and con. One moment I thought it would prove a miserable failure, and Bartleby would be found all alive at my office as usual; the next moment it seemed certain that I should see his chair empty. And so I kept veering about. At the corner of Broadway and Canal- street, I saw quite an excited group of people standing in earnest conversation.

+

"I'll take odds he doesn't," said a voice as I passed.

+

"Doesn't go?--done!" said I, "put up your money."

+

I was instinctively putting my hand in my pocket to produce my own, when I remembered that this was an election day. The words I had overheard bore no reference to Bartleby, but to the success or non-success of some candidate for the mayoralty. In my intent frame of mind, I had, as it were, imagined that all Broadway shared in my excitement, and were debating the same question with me. I passed on, very thankful that the uproar of the street screened my momentary absent-mindedness.

+

As I had intended, I was earlier than usual at my office door. I stood listening for a moment. All was still. He must be gone. I tried the knob. The door was locked. Yes, my procedure had worked to a charm; he indeed must be vanished. Yet a certain melancholy mixed with this: I was almost sorry for my brilliant success. I was fumbling under the door mat for the key, which Bartleby was to have left there for me, when accidentally my knee knocked against a panel, producing a summoning sound, and in response a voice came to me from within--"Not yet; I am occupied."

+

It was Bartleby.

+

I was thunderstruck. For an instant I stood like the man who, pipe in mouth, was killed one cloudless afternoon long ago in Virginia, by summer lightning; at his own warm open window he was killed, and remained leaning out there upon the dreamy afternoon, till some one touched him, when he fell. "Not gone!" I murmured at last. But again obeying that wondrous ascendancy which the inscrutable scrivener had over me, and from which ascendancy, for all my chafing, I could not completely escape, I slowly went down stairs and out into the street, and while walking round the block, considered what I should next do in this unheard-of-perplexity. Turn the man out by an actual thrusting I could not; to drive him away by calling him hard names would not do; calling in the police was an unpleasant idea; and yet, permit him to enjoy his cadaverous triumph over me,--this too I could not think of. What was to be done? or, if nothing could be done, was there any thing further that I could assume in the matter? Yes, as before I had prospectively assumed that Bartleby would depart, so now I might retrospectively assume that departed he was. In the legitimate carrying out of this assumption, I might enter my office in a great hurry, and pretending not to see Bartleby at all, walk straight against him as if he were air. Such a proceeding would in a singular degree have the appearance of a home-thrust. It was hardly possible that Bartleby could withstand such an application of the doctrine of assumptions. But upon second thoughts the success of the plan seemed rather dubious. I resolved to argue the matter over with him again.

+

Bartleby," said I, entering the office, with a quietly severe expression. "I am seriously displeased. I am pained, Bartleby. I had thought better of you. I had imagined you of such a gentlemanly organization, that in any delicate dilemma a slight hint would suffice--in short, an assumption. But it appears I am deceived. Why," I added, unaffectedly starting, "you have not even touched the money yet," pointing to it, just where I had left it the evening previous.

+

He answered nothing.

+

"Will you, or will you not, quit me?" I now demanded in a sudden passion, advancing close to him.

+

"I would prefer not to quit you," he replied, gently emphasizing the not.

+

"What earthly right have you to stay here? do you pay any rent? Do you pay my taxes? Or is this property yours?"

+

He answered nothing.

+

"Are you ready to go on and write now? Are your eyes recovered? Could you copy a small paper for me this morning? or help examine a few lines? or step round to the post-office? In a word, will you do any thing at all, to give a coloring to your refusal to depart the premises?"

+

He silently retired into his hermitage.

+

I was now in such a state of nervous resentment that I thought it but prudentto check myself at present from further demonstrations. Bartleby and I were alone. I remembered the tragedy of the unfortunate Adams and the still more unfortunate Colt in the solitary office of the latter; and how poor Colt, being dreadfully incensed by Adams, and imprudently permitting himself to get wildly excited, was at unawares hurried into his fatal act--an act which certainly no man could possibly deplore more than the actor himself. Often it had occurred to me in my ponderings upon the subject, that had that altercation taken place in the public street, or at a private residence, it would not have terminated as it did. It was the circumstance of being alone in a solitary office, up stairs, of a building entirely unhallowed by humanizing domestic associations--an uncarpeted office, doubtless of a dusty, haggard sort of appearance;--this it must have been, which greatly helped to enhance the irritable desperation of the hapless Colt.

+

But when this old Adam of resentment rose in me and tempted me concerning Bartleby, I grappled him and threw him. How? Why, simply by recalling the divine injunction: "A new commandment give I unto you, that ye love one another." Yes, this it was that saved me. Aside from higher considerations, charity often operates as a vastly wise and prudent principle--a great safeguard to its possessor. Men have committed murder for jealousy's sake, and anger's sake, and hatred's sake, and selfishness' sake, and spiritual pride's sake; but no man that ever I heard of, ever committed a diabolical murder for sweet charity's sake. Mere self-interest, then, if no better motive can be enlisted, should, especially with high-tempered men, prompt all beings to charity and philanthropy. At any rate, upon the occasion in question, I strove to drown my exasperated feelings towards the scrivener by benevolently construing his conduct. Poor fellow, poor fellow! thought I, he don't mean any thing; and besides, he has seen hard times, and ought to be indulged.

+

I endeavored also immediately to occupy myself, and at the same time to comfort my despondency.I tried to fancy that in the course of the morning, at such time as might prove agreeable to him, Bartleby, of his own free accord, would emerge from his hermitage, and take up some decided line of march in the direction of the door. But no. Half-past twelve o'clock came; Turkey began to glow in the face, overturn his inkstand, and become generally obstreperous; Nippers abated down into quietude and courtesy; Ginger Nut munched his noon apple; and Bartleby remained standing at his window in one of his profoundest deadwall reveries. Will it be credited? Ought I to acknowledge it? That afternoon I left the office without saying one further word to him.

+

Some days now passed, during which, at leisure intervals I looked a little into Edwards on the Will," and "Priestly on Necessity." Under the circumstances, those books induced a salutary feeling. Gradually I slid into the persuasion that these troubles of mine touching the scrivener, had been all predestinated from eternity, and Bartleby was billeted upon me for some mysterious purpose of an all-wise Providence, which it was not for a mere mortal like me to fathom. Yes, Bartleby, stay there behind your screen, thought I; I shall persecute you no more; you are harmless and noiseless as any of these old chairs; in short, I never feel so private as when I know you are here. At least I see it, I feel it; I penetrate to the predestinated purpose of my life. I am content. Others may have loftier parts to enact; but my mission in this world, Bartleby, is to furnish you with office-room for such period as you may see fit to remain.

+

I believe that this wise and blessed frame of mind would have continued with me, had it not been for the unsolicited and uncharitable remarks obtruded upon me by my professional friends who visited the rooms. But thus it often is, that the constant friction of illiberal minds wears out at last the best resolves of the more generous. Though to be sure, when I reflected upon it, it was not strange that people entering my office should be struck by the peculiar aspect of the unaccountable Bartleby, and so be tempted to throw out some sinister observations concerning him. Sometimes an attorney having business with me, and calling at my office, and finding no one but the scrivener there, would undertake to obtain some sort of precise information from him touching my whereabouts; but without heeding his idle talk, Bartleby would remain standing immovable in the middle of the room. So after contemplating him in that position for a time, the attorney would depart, no wiser than he came.

+

Also, when a Reference was going on, and the room full of lawyers and witnesses and business was driving fast; some deeply occupied legal gentleman present, seeing Bartleby wholly unemployed, would request him to run round to his (the legal gentleman's) office and fetch some papers for him. Thereupon, Bartleby would tranquilly decline, and remain idle as before. Then the lawyer would give a great stare, and turn to me. And what could I say? At last I was made aware that all through the circle of my professional acquaintance, a whisper of wonder was running round, having reference to the strange creature I kept at my office. This worried me very much. And as the idea came upon me of his possibly turning out a long-lived man, and keep occupying my chambers, and denying my authority; and perplexing my visitors; and scandalizing my professional reputation; and casting a general gloom over the premises; keeping soul and body together to the last upon his savings (for doubtless he spent but half a dime a day), and in the end perhaps outlive me, and claim possession of my office by right of his perpetual occupancy: as all these dark anticipations crowded upon me more and more, and my friends continually intruded their relentless remarks upon the apparition in my room; a great change was wrought in me. I resolved to gather all my faculties together, and for ever rid me of this intolerable incubus.

+

Ere revolving any complicated project, however, adapted to this end, I first simply suggested to Bartleby the propriety of his permanent departure. In a calm and serious tone, I commended the idea to his careful and mature consideration. But having taken three days to meditate upon it, he apprised me that his original determination remained the same; in short, that he still preferred to abide with me.

+

What shall I do? I now said to myself, buttoning up my coat to the last button. What shall I do? what ought I to do? what does conscience say I should do with this man, or rather ghost. Rid myself of him, I must; go, he shall. But how? You will not thrust him, the poor, pale, passive mortal,--you will not thrust such a helpless creature out of your door? you will not dishonor yourself by such cruelty? No, I will not, I cannot do that. Rather would I let him live and die here, and then mason up his remains in the wall. What then will you do? For all your coaxing, he will not budge. Bribes he leaves under your own paperweight on your table; in short, it is quite plain that he prefers to cling to you.

+

Then something severe, something unusual must be done. What! surely you will not have him collared by a constable, and commit his innocent pallor to the common jail? And upon what ground could you procure such a thing to be done?--a vagrant, is he? What! he a vagrant, a wanderer, who refuses to budge? It is because he will not be a vagrant, then, that you seek to count him as a vagrant. That is too absurd. No visible means of support: there I have him. Wrong again: for indubitably he does support himself, and that is the only unanswerable proof that any man can show of his possessing the means so to do. No more then. Since he will not quit me, I must quit him. I will change my offices; I will move elsewhere; and give him fair notice, that if I find him on my new premises I will then proceed against him as a common trespasser.

+

Acting accordingly, next day I thus addressed him: "I find these chambers too far from the City Hall; the air is unwholesome. In a word, I propose to remove my offices next week, and shall no longer require your services. I tell you this now, in order that you may seek another place."

+

He made no reply, and nothing more was said.

+

On the appointed day I engaged carts and men, proceeded to my chambers, and having but little furniture, every thing was removed in a few hours. Throughout, the scrivener remained standing behind the screen, which I directed to be removed the last thing. It was withdrawn; and being folded up like a huge folio, left him the motionless occupant of a naked room. I stood in the entry watching him a moment, while something from within me upbraided me.

+

I re-entered, with my hand in my pocket--and--and my heart in my mouth.

+

"Good-bye, Bartleby; I am going--good-bye, and God some way bless you; and take that," slipping something in his hand. But it dropped to the floor, and then,--strange to say--I tore myself from him whom I had so longed to be rid of.

+

Established in my new quarters, for a day or two I kept the door locked, and started at every footfall in the passages. When I returned to my rooms after any little absence, I would pause at the threshold for an instant, and attentively listen, ere applying my key. But these fears were needless. Bartleby never came nigh me.

+

I thought all was going well, when a perturbed looking stranger visited me, inquiring whether I was the person who had recently occupied rooms at No.--Wall-street.

+

Full of forebodings, I replied that I was.

+

"Then, sir," said the stranger, who proved a lawyer, "you are responsible for the man you left there. He refuses to do any copying; he refuses to do any thing; he says he prefers not to; and he refuses to quit the premises."

+

"I am very sorry, sir," said I, with assumed tranquillity, but an inward tremor, "but, really, the man you allude to is nothing to me --he is no relation or apprentice of mine, that you should hold me responsible for him."

+

"In mercy's name, who is he?"

+

"I certainly cannot inform you. I know nothing about him. Formerly I employed him as a copyist; but he has done nothing for me now for some time past."

+

"I shall settle him then,--good morning, sir."

+

Several days passed, and I heard nothing more; and though I often felt a charitable prompting to call at the place and see poor Bartleby, yet a certain squeamishness of I know not what withheld me.

+

All is over with him, by this time, thought I at last, when through another week no further intelligence reached me. But coming to my room the day after, I found several persons waiting at my door in a high state of nervous excitement.

+

"That's the man--here he comes," cried the foremost one, whom recognized as the lawyer who had previously called upon me alone.

+

"You must take him away, sir, at once," cried a portly person among them, advancing upon me, and whom I knew to be the landlord of No.--Wall-street. "These gentlemen, my tenants, cannot stand it any longer; Mr. B--" pointing to the lawyer, "has turned him out of his room, and he now persists in haunting the buildinggenerally, sitting upon the banisters of the stairs by day, and sleeping in the entry by night. Every body is concerned; clients are leaving the offices; some fears are entertained of a mob; something you must do, and that without delay."

+

Aghast at this torment, I fell back before it, and would fain have locked myselfin my new quarters. In vain I persisted that Bartleby was nothing to me--no more than to any one else. In vain:--I was the last person known to have any thing to do with him, and they held me to the terrible account. Fearful then of being exposed in the papers (as one person present obscurely threatened) I considered the matter, and at length said, that if the lawyer would give me a confidential interview with the scrivener, in his (the lawyer's) own room, I would that afternoon strive my best to rid them of the nuisance they complained of.

+

Going up stairs to my old haunt, there was Bartleby silently sitting upon the banister at the landing.

+

"What are you doing here, Bartleby?" said I.

+

"Sitting upon the banister," he mildly replied.

+

I motioned him into the lawyer's room, who then left us.

+

"Bartleby," said I, "are you aware that you are the cause of great tribulation to me, by persisting in occupying the entry after being dismissed from the office?"

+

No answer.

+

"Now one of two things must take place. Either you must do something or something must be done to you. Now what sort of business would you like to engage in? Would you like to re-engage in copying for some one?"

+

"No; I would prefer not to make any change."

+

"Would you like a clerkship in a dry-goods store?"

+

"There is too much confinement about that. No, I would not like a clerkship; but I am not particular."

+

"Too much confinement," I cried, "why you keep yourself confined all the time!"

+

"I would prefer not to take a clerkship," he rejoined, as if to settle that little item at once.

+

"How would a bar-tender's business suit you? There is no trying of the eyesight in that."

+

"I would not like it at all; though, as I said before, I am not particular."

+

His unwonted wordiness inspirited me. I returned to the charge.

+

"Well then, would you like to travel through the country collecting bills for the merchants? That would improve your health."

+

"No, I would prefer to be doing something else."

+

"How then would going as a companion to Europe, to entertain some young gentleman with your conversation,--how would that suit you?"

+

"Not at all. It does not strike me that there is any thing definite about that. I like to be stationary. But I am not particular.

+

"Stationary you shall be then," I cried, now losing all patience, and for the first time in all my exasperating connection with him fairly flying into a passion. "If you do not go away from these premises before night, I shall feel bound--indeed I am bound--to-- to--to quit the premises myself!" I rather absurdly concluded, knowing not with what possible threat to try to frighten his immobility into compliance. Despairing of all further efforts, I was precipitately leaving him, when a final thought occurred to me--one which had not been wholly unindulged before.

+

"Bartleby," said I, in the kindest tone I could assume under such exciting circumstances, "will you go home with me now--not to my office, but my dwelling--and remain there till we can conclude upon some convenient arrangement for you at our leisure? Come, let us start now, right away."

+

"No: at present I would prefer not to make any change at all."

+

I answered nothing; but effectualy dodging every one by the suddenness and rapidity of my flight, rushed from the building, ran up Wall-street towards Broadway, and jumping into the first omnibus was soon removed from pursuit. As soon as tranquility returned I distinctly perceived that I had now done all that I possibly could, both in respect to the demands of the landlord and his tenants, and with regard to my own desire and sense of duty, to benefit Bartleby, and shield him from rude persecution. I now strove to be entirely care-free and quiescent; and my conscience justified me in the attempt; though indeed it was not so successful as I could have wished. So fearful was I of being again hunted out by the incensed landlord and his exasperated tenants, that, surrendering my business to Nippers, for a few days I drove about the upper part of the town and through the suburbs, in my rockaway; crossed over to Jersey City and Hoboken, and paid fugitive visits to Manhattanville and Astoria. In fact I almost lived in my rockaway for the time.

+

When again I entered my office, lo, a note from the landlord lay upon desk. opened it with trembling hands. informed me that writer had sent to police, and Bartleby removed the Tombs as a vagrant. Moreover, since I knew more about him than any one else, he wished me to appear at that place, and make a suitable statement of the facts. These tidings had a conflicting effect upon me. At first I was indignant; but at last almost approved. The landlord's energetic, summary disposition, had led him to adopt a procedure which I do not think I would have decided upon myself; and yet as a last resort, under such peculiar circumstances, it seemed the only plan.

+

As I afterwards learned, the poor scrivener, when told that he must be conducted to the Tombs, offered not the slightest obstacle, but in his pale unmoving way, silently acquiesced.

+

Some of the compassionate and curious bystanders joined the party; and headed by one of the constables arm in arm with Bartleby, the silent procession filed its way through all the noise, and heat, and joy of the roaring thoroughfares at noon.

+

The same day I received the note I went to the Tombs, or to speak more properly, the Halls of Justice. Seeking the right officer, I stated the purpose of my call, and was informed that the individual I described was indeed within. I then assured the functionary that Bartleby was a perfectly honest man, and greatly to be compassionated, however unaccountably eccentric. I narrated all I knew,and closed by suggesting the idea of letting him remain in as indulgent confinement as possible till something less harsh might be done--though indeed I hardly knew what. At all events, if nothing else could be decided upon, the alms-house must receive him. I then begged to have an interview.

+

Being under no disgraceful charge, and quite serene and harmless in all his ways, they had permitted him freely to wander about the prison, and especially in the inclosed grass-platted yards thereof. And so I found him there, standing all alone in the quietest of the yards, his face towards a high wall, while all around, from the narrow slits of the jail windows, I thought I saw peering out upon him the eyes of murderers and thieves.

+

"Bartleby!"

+

"I know you," he said, without looking round,--"and I want nothing to say to you."

+

"It was not I that brought you here, Bartleby," said I, keenly pained at his implied suspicion. "And to you, this should not be so vile a place. Nothing reproachful attaches to you by being here. And see, it is not so sad a place as one might think. Look, there is the sky, and here is the grass."

+

"I know where I am," he replied, but would say nothing more, and so I left him.

+

As I entered the corridor again, a broad meat-like man in an apron, accosted me, and jerking his thumb over his shoulder said--"Is that your friend?"

+

"Yes."

+

"Does he want to starve? If he does, let him live on the prison fare, that's all.

+

"Who are you?" asked I, not knowing what to make of such an unofficially speaking person in such a place.

+

"I am the grub-man. Such gentlemen as have friends here, hire me to provide them with something good to eat."

+

"Is this so?" said I, turning to the turnkey.

+

He said it was.

+

"Well then," said I, slipping some silver into the grub-man's hands (for so they called him). "I want you to give particular attention to my friend there; let him have the best dinner you can get. And you must be as polite to him as possible."

+

"Introduce me, will you?" said the grub-man, looking at me with an expression which seemed to say he was all impatience for an opportunity to give a specimen of his breeding.

+

Thinking it would prove of benefit to the scrivener, I acquiesced; and asking the grub-man his name, went up with him to Bartleby.

+

"Bartleby, this is a friend; you will find him very useful to you."

+

"Your sarvant, sir, your sarvant," said the grub-man, making a low salutation behind his apron. "Hope you find it pleasant here, sir;--spacious grounds--cool apartments, sir--hope you'll stay with us some time--try to make it agreeable. What will you have for dinner today?"

+

"I prefer not to dine to-day," said Bartleby, turning away. "It would disagree with me; I am unused to dinners." So saying he slowly moved to the other side of the inclosure, and took up a position fronting the dead-wall.

+

"How's this?" said the grub-man, addressing me with a stare of astonishment. "He's odd, aint he?"

+

"I think he is a little deranged," said I, sadly.

+

"Deranged? deranged is it? Well now, upon my word, I thought that friend of yourn was a gentleman forger; they are always pale and genteel-like, them forgers. I can't help pity 'em--can't help it, sir. Did you know Monroe Edwards?" he added touchingly, and paused. Then, laying his hand pityingly on my shoulder, sighed, "he died of consumption at Sing-Sing. so you weren't acquainted with Monroe?"

+

"No, I was never socially acquainted with any forgers. But I cannot stop longer. Look to my friend yonder. You will not lose by it. I will see you again."

+

Some few days after this, I again obtained admission to the Tombs, and went through the corridors in quest of Bartleby; but without finding him.

+

"I saw him coming from his cell not long ago," said a turnkey, "may be he's gone to loiter in the yards."

+

So I went in that direction.

+

"Are you looking for the silent man?" said another turnkey passing me. "Yonder he lies--sleeping in the yard there. 'Tis not twenty minutes since I saw him lie down."

+

The yard was entirely quiet. It was not accessible to the common prisoners. The surrounding walls, of amazing thickness, kept off all sound behind them. The Egyptian character of the masonry weighed upon me with its gloom. But a soft imprisoned turf grew under foot. The heart of the eternal pyramids, it seemed, wherein, by some strange magic, through the clefts, grass-seed, dropped by birds, had sprung.

+

Strangely huddled at the base of the wall, his knees drawn up, and lying on his side, his head touching the cold stones, I saw the wasted Bartleby. But nothing stirred. I paused; then went close up to him; stooped over, and saw that his dim eyes were open; otherwise he seemed profoundly sleeping. Something prompted me to touch him. I felt his hand, when a tingling shiver ran up my arm and down my spine to my feet.

+

The round face of the grub-man peered upon me now. "His dinner is ready. Won't he dine to-day, either? Or does he live without dining?"

+

"Lives without dining," said I, and closed the eyes.

+

"Eh!--He's asleep, aint he?"

+

"With kings and counsellors," murmured I.

+

* * * * * * * *

+

There would seem little need for proceeding further in this history. Imagination will readily supply the meagre recital of poor Bartleby's interment. But ere parting with the reader, let me say, that if this little narrative has sufficiently interested him, to awaken curiosity as to who Bartleby was, and what manner of life he led prior to the present narrator's making his acquaintance, I can only reply, that in such curiosity I fully share, but am wholly unable to gratify it. Yet here I hardly know whether I should divulge one little item of rumor, which came to my ear a few months after the scrivener's decease. Upon what basis it rested, I could never ascertain; and hence how true it is I cannot now tell. But inasmuch as this vague report has not been without a certain strange suggestive interest to me, however said, it may prove the same with some others; and so I will briefly mention it. The report was this: that Bartleby had been a subordinate clerk in the Dead Letter Office at Washington, from which he had been suddenly removed by a change in the administration. When I think over this rumor, I cannot adequately express the emotions which seize me. Dead letters! does it not sound like dead men? Conceive a man by nature and misfortune prone to a pallid hopelessness, can any business seem more fitted to heighten it than that of continually handling these dead letters and assorting them for the flames? For by the cart-load they are annually burned. Sometimes from out the folded paper the pale clerk takes a ring:--the bank-note sent in swiftest charity:--he whom it would relieve, nor eats nor hungers any more; pardon for those who died despairing; hope for those who died unhoping; good tidings for those who died stifled by unrelieved calamities. On errands of life, these letters speed to death.

+

Ah Bartleby! Ah humanity!

\ No newline at end of file diff --git a/test/test-pages/http2-spec/expected-metadata.json b/test/test-pages/http2-spec/expected-metadata.json new file mode 100644 index 0000000..59424d1 --- /dev/null +++ b/test/test-pages/http2-spec/expected-metadata.json @@ -0,0 +1,6 @@ +{ + "title": "Hypertext Transfer Protocol version 2", + "byline": "Authors' Addresses", + "excerpt": "This specification describes an optimized expression of the semantics of the Hypertext Transfer Protocol (HTTP). HTTP/2 enables a more efficient use of network resources and a reduced perception of latency by introducing header field compression and allowing multiple concurrent exchanges on the same connection. It also introduces unsolicited push of representations from servers to clients. This specification is an alternative to, but does not obsolete, the HTTP/1.1 message syntax. HTTP's existing semantics remain unchanged.", + "readerable": true +} diff --git a/test/test-pages/http2-spec/expected.html b/test/test-pages/http2-spec/expected.html new file mode 100644 index 0000000..8903f85 --- /dev/null +++ b/test/test-pages/http2-spec/expected.html @@ -0,0 +1,1421 @@ +
+
+

Hypertext Transfer Protocol version 2 +
draft-ietf-httpbis-http2-latest

+

This specification describes an optimized expression of the semantics of the Hypertext Transfer Protocol (HTTP). HTTP/2 enables a more efficient use of network resources and a reduced perception of latency by introducing header field compression and allowing multiple concurrent exchanges on the same connection. It also introduces unsolicited push of representations from servers to clients.

+

This specification is an alternative to, but does not obsolete, the HTTP/1.1 message syntax. HTTP's existing semantics remain unchanged.

+

Discussion of this draft takes place on the HTTPBIS working group mailing list (ietf-http-wg@w3.org), which is archived at <https://lists.w3.org/Archives/Public/ietf-http-wg/>.

+

Working Group information can be found at <https://tools.ietf.org/wg/httpbis/>; that specific to HTTP/2 are at <https://http2.github.io/>.

+

The changes in this draft are summarized in Appendix B.

+
+

This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79.

+

Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet-Drafts is at http://datatracker.ietf.org/drafts/current/.

+

Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as “work in progress”.

+

This Internet-Draft will expire on November 5, 2015.

+
+
+

Copyright © 2015 IETF Trust and the persons identified as the document authors. All rights reserved.

+

This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (http://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Code Components extracted from this document must include Simplified BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Simplified BSD License.

+
+
+ +
+

The Hypertext Transfer Protocol (HTTP) is a wildly successful protocol. However, how HTTP/1.1 uses the underlying transport ( + [RFC7230] , Section 6) has several characteristics that have a negative overall effect on application performance today.

+

In particular, HTTP/1.0 allowed only one request to be outstanding at a time on a given TCP connection. HTTP/1.1 added request pipelining, but this only partially addressed request concurrency and still suffers from head-of-line blocking. Therefore, HTTP/1.0 and HTTP/1.1 clients that need to make many requests use multiple connections to a server in order to achieve concurrency and thereby reduce latency.

+

Furthermore, HTTP header fields are often repetitive and verbose, causing unnecessary network traffic, as well as causing the initial TCP [TCP] congestion window to quickly fill. This can result in excessive latency when multiple requests are made on a new TCP connection.

+

HTTP/2 addresses these issues by defining an optimized mapping of HTTP's semantics to an underlying connection. Specifically, it allows interleaving of request and response messages on the same connection and uses an efficient coding for HTTP header fields. It also allows prioritization of requests, letting more important requests complete more quickly, further improving performance.

+

The resulting protocol is more friendly to the network, because fewer TCP connections can be used in comparison to HTTP/1.x. This means less competition with other flows, and longer-lived connections, which in turn leads to better utilization of available network capacity.

+

Finally, HTTP/2 also enables more efficient processing of messages through use of binary message framing.

+
+
+

HTTP/2 provides an optimized transport for HTTP semantics. HTTP/2 supports all of the core features of HTTP/1.1, but aims to be more efficient in several ways.

+

The basic protocol unit in HTTP/2 is a frame (Section 4.1). Each frame type serves a different purpose. For example, HEADERS and DATA frames form the basis of HTTP requests and responses (Section 8.1); other frame types like SETTINGS, WINDOW_UPDATE, and PUSH_PROMISE are used in support of other HTTP/2 features.

+

Multiplexing of requests is achieved by having each HTTP request-response exchange associated with its own stream (Section 5). Streams are largely independent of each other, so a blocked or stalled request or response does not prevent progress on other streams.

+

Flow control and prioritization ensure that it is possible to efficiently use multiplexed streams. Flow control (Section 5.2) helps to ensure that only data that can be used by a receiver is transmitted. Prioritization (Section 5.3) ensures that limited resources can be directed to the most important streams first.

+

HTTP/2 adds a new interaction mode, whereby a server can push responses to a client (Section 8.2). Server push allows a server to speculatively send data to a client that the server anticipates the client will need, trading off some network usage against a potential latency gain. The server does this by synthesizing a request, which it sends as a PUSH_PROMISE frame. The server is then able to send a response to the synthetic request on a separate stream.

+

Because HTTP header fields used in a connection can contain large amounts of redundant data, frames that contain them are compressed (Section 4.3). This has especially advantageous impact upon request sizes in the common case, allowing many requests to be compressed into one packet.

+
+

2.1 Document Organization +

+
+

The HTTP/2 specification is split into four parts:

+
    +
  • Starting HTTP/2 (Section 3) covers how an HTTP/2 connection is initiated.
  • +
  • The framing (Section 4) and streams (Section 5) layers describe the way HTTP/2 frames are structured and formed into multiplexed streams.
  • +
  • Frame (Section 6) and error (Section 7) definitions include details of the frame and error types used in HTTP/2.
  • +
  • HTTP mappings (Section 8) and additional requirements (Section 9) describe how HTTP semantics are expressed using frames and streams.
  • +
+
+

While some of the frame and stream layer concepts are isolated from HTTP, this specification does not define a completely generic framing layer. The framing and streams layers are tailored to the needs of the HTTP protocol and server push.

+
+
+

2.2 Conventions and Terminology +

+

The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in RFC 2119 [RFC2119].

+

All numeric values are in network byte order. Values are unsigned unless otherwise indicated. Literal values are provided in decimal or hexadecimal as appropriate. Hexadecimal literals are prefixed with 0x to distinguish them from decimal literals.

+
+

The following terms are used:

+
client:
+
The endpoint that initiates an HTTP/2 connection. Clients send HTTP requests and receive HTTP responses.
connection:
+
A transport-layer connection between two endpoints.
connection error:
+
An error that affects the entire HTTP/2 connection.
endpoint:
+
Either the client or server of the connection.
frame:
+
The smallest unit of communication within an HTTP/2 connection, consisting of a header and a variable-length sequence of octets structured according to the frame type.
peer:
+
An endpoint. When discussing a particular endpoint, "peer" refers to the endpoint that is remote to the primary subject of discussion.
receiver:
+
An endpoint that is receiving frames.
sender:
+
An endpoint that is transmitting frames.
server:
+
The endpoint that accepts an HTTP/2 connection. Servers receive HTTP requests and serve HTTP responses.
stream:
+
A bi-directional flow of frames within the HTTP/2 connection.
stream error:
+
An error on the individual HTTP/2 stream.
+
+
+

Finally, the terms "gateway", "intermediary", "proxy", and "tunnel" are defined in Section 2.3 of + [RFC7230] . Intermediaries act as both client and server at different times.

+

The term "payload body" is defined in Section 3.3 of + [RFC7230] .

+
+
+
+

An HTTP/2 connection is an application layer protocol running on top of a TCP connection ( + [TCP] ). The client is the TCP connection initiator.

+

HTTP/2 uses the same "http" and "https" URI schemes used by HTTP/1.1. HTTP/2 shares the same default port numbers: 80 for "http" URIs and 443 for "https" URIs. As a result, implementations processing requests for target resource URIs like http://example.org/foo or https://example.com/bar are required to first discover whether the upstream server (the immediate peer to which the client wishes to establish a connection) supports HTTP/2.

+

The means by which support for HTTP/2 is determined is different for "http" and "https" URIs. Discovery for "http" URIs is described in Section 3.2. Discovery for "https" URIs is described in Section 3.3.

+
+

3.1 HTTP/2 Version Identification

+
+

The protocol defined in this document has two identifiers.

+
    +
  • +

    The string "h2" identifies the protocol where HTTP/2 uses TLS [TLS12]. This identifier is used in the TLS application layer protocol negotiation extension (ALPN) [TLS-ALPN] field and in any place where HTTP/2 over TLS is identified.

    +

    The "h2" string is serialized into an ALPN protocol identifier as the two octet sequence: 0x68, 0x32.

    +
  • +
  • +

    The string "h2c" identifies the protocol where HTTP/2 is run over cleartext TCP. This identifier is used in the HTTP/1.1 Upgrade header field and in any place where HTTP/2 over TCP is identified.

    +

    The "h2c" string is reserved from the ALPN identifier space, but describes a protocol that does not use TLS.

    +
  • +
+
+

Negotiating "h2" or "h2c" implies the use of the transport, security, framing and message semantics described in this document.

+

Only implementations of the final, published RFC can identify themselves as "h2" or "h2c". Until such an RFC exists, implementations MUST NOT identify themselves using these strings.

+

Implementations of draft versions of the protocol MUST add the string "-" and the corresponding draft number to the identifier. For example, draft-ietf-httpbis-http2-11 over TLS is identified using the string "h2-11".

+

Non-compatible experiments that are based on these draft versions MUST append the string "-" and an experiment name to the identifier. For example, an experimental implementation of packet mood-based encoding based on draft-ietf-httpbis-http2-09 might identify itself as "h2-09-emo". Note that any label MUST conform to the "token" syntax defined in Section 3.2.6 of + [RFC7230] . Experimenters are encouraged to coordinate their experiments on the ietf-http-wg@w3.org mailing list.

+
+
+

3.2 Starting HTTP/2 for "http" URIs

+

A client that makes a request for an "http" URI without prior knowledge about support for HTTP/2 on the next hop uses the HTTP Upgrade mechanism (Section 6.7 of + [RFC7230] ). The client does so by making an HTTP/1.1 request that includes an Upgrade header field with the "h2c" token. Such an HTTP/1.1 request MUST include exactly one HTTP2-Settings (Section 3.2.1) header field.

+
+

For example:

GET / HTTP/1.1
+Host: server.example.com
+Connection: Upgrade, HTTP2-Settings
+Upgrade: h2c
+HTTP2-Settings: <base64url encoding of HTTP/2 SETTINGS payload>
+
+
+

Requests that contain an payload body MUST be sent in their entirety before the client can send HTTP/2 frames. This means that a large request can block the use of the connection until it is completely sent.

+

If concurrency of an initial request with subsequent requests is important, an OPTIONS request can be used to perform the upgrade to HTTP/2, at the cost of an additional round-trip.

+

A server that does not support HTTP/2 can respond to the request as though the Upgrade header field were absent:

+
HTTP/1.1 200 OK
+Content-Length: 243
+Content-Type: text/html
+
+...
+
+

A server MUST ignore an "h2" token in an Upgrade header field. Presence of a token with "h2" implies HTTP/2 over TLS, which is instead negotiated as described in Section 3.3.

+

A server that supports HTTP/2 accepts the upgrade with a 101 (Switching Protocols) response. After the empty line that terminates the 101 response, the server can begin sending HTTP/2 frames. These frames MUST include a response to the request that initiated the Upgrade.

+
+

For example:

HTTP/1.1 101 Switching Protocols
+Connection: Upgrade
+Upgrade: h2c
+
+[ HTTP/2 connection ...
+
+

The first HTTP/2 frame sent by the server MUST be a SETTINGS frame (Section 6.5) as the server connection preface (Section 3.5). Upon receiving the 101 response, the client MUST send a connection preface (Section 3.5), which includes a SETTINGS frame.

+

The HTTP/1.1 request that is sent prior to upgrade is assigned a stream identifier of 1 (see Section 5.1.1) with default priority values (Section 5.3.5). Stream 1 is implicitly "half closed" from the client toward the server (see Section 5.1), since the request is completed as an HTTP/1.1 request. After commencing the HTTP/2 connection, stream 1 is used for the response.

+
+
+

3.4 Starting HTTP/2 with Prior Knowledge

+

A client can learn that a particular server supports HTTP/2 by other means. For example, + [ALT-SVC] describes a mechanism for advertising this capability.

+

A client MUST send the connection preface (Section 3.5), and then MAY immediately send HTTP/2 frames to such a server; servers can identify these connections by the presence of the connection preface. This only affects the establishment of HTTP/2 connections over cleartext TCP; implementations that support HTTP/2 over TLS MUST use protocol negotiation in TLS [TLS-ALPN].

+

Likewise, the server MUST send a connection preface (Section 3.5).

+

Without additional information, prior support for HTTP/2 is not a strong signal that a given server will support HTTP/2 for future connections. For example, it is possible for server configurations to change, for configurations to differ between instances in clustered servers, or for network conditions to change.

+
+
+
+

Once the HTTP/2 connection is established, endpoints can begin exchanging frames.

+
+

4.2 Frame Size

+

The size of a frame payload is limited by the maximum size that a receiver advertises in the SETTINGS_MAX_FRAME_SIZE setting. This setting can have any value between 214 (16,384) and 224-1 (16,777,215) octets, inclusive.

+
+

All implementations MUST be capable of receiving and minimally processing frames up to 214 octets in length, plus the 9 octet frame header (Section 4.1). The size of the frame header is not included when describing frame sizes.

+
Note:
+
Certain frame types, such as PING (Section 6.7), impose additional limits on the amount of payload data allowed.
+
+
+

An endpoint MUST send a FRAME_SIZE_ERROR error if a frame exceeds the size defined in SETTINGS_MAX_FRAME_SIZE, any limit defined for the frame type, or it is too small to contain mandatory frame data. A frame size error in a frame that could alter the state of the entire connection MUST be treated as a connection error (Section 5.4.1); this includes any frame carrying a header block (Section 4.3) (that is, HEADERS, PUSH_PROMISE, and CONTINUATION), SETTINGS, and any frame with a stream identifier of 0.

+

Endpoints are not obligated to use all available space in a frame. Responsiveness can be improved by using frames that are smaller than the permitted maximum size. Sending large frames can result in delays in sending time-sensitive frames (such as RST_STREAM, WINDOW_UPDATE, or PRIORITY) which if blocked by the transmission of a large frame, could affect performance.

+
+
+
+
+

A "stream" is an independent, bi-directional sequence of frames exchanged between the client and server within an HTTP/2 connection. Streams have several important characteristics:

+
    +
  • A single HTTP/2 connection can contain multiple concurrently open streams, with either endpoint interleaving frames from multiple streams.
  • +
  • Streams can be established and used unilaterally or shared by either the client or server.
  • +
  • Streams can be closed by either endpoint.
  • +
  • The order in which frames are sent on a stream is significant. Recipients process frames in the order they are received. In particular, the order of HEADERS, and DATA frames is semantically significant.
  • +
  • Streams are identified by an integer. Stream identifiers are assigned to streams by the endpoint initiating the stream.
  • +
+
+
+

5.1 Stream States

+

The lifecycle of a stream is shown in Figure 2.

+

Figure 2: Stream States

+

Note that this diagram shows stream state transitions and the frames and flags that affect those transitions only. In this regard, CONTINUATION frames do not result in state transitions; they are effectively part of the HEADERS or PUSH_PROMISE that they follow. For the purpose of state transitions, the END_STREAM flag is processed as a separate event to the frame that bears it; a HEADERS frame with the END_STREAM flag set can cause two state transitions.

+

Both endpoints have a subjective view of the state of a stream that could be different when frames are in transit. Endpoints do not coordinate the creation of streams; they are created unilaterally by either endpoint. The negative consequences of a mismatch in states are limited to the "closed" state after sending RST_STREAM, where frames might be received for some time after closing.

+
+

Streams have the following states:

+
idle:
+
+

+
All streams start in the "idle" state.

+

The following transitions are valid from this state:

+
    +
  • Sending or receiving a HEADERS frame causes the stream to become "open". The stream identifier is selected as described in Section 5.1.1. The same HEADERS frame can also cause a stream to immediately become "half closed".
  • +
  • Sending a PUSH_PROMISE frame on another stream reserves the idle stream that is identified for later use. The stream state for the reserved stream transitions to "reserved (local)".
  • +
  • Receiving a PUSH_PROMISE frame on another stream reserves an idle stream that is identified for later use. The stream state for the reserved stream transitions to "reserved (remote)".
  • +
  • Note that the PUSH_PROMISE frame is not sent on the idle stream, but references the newly reserved stream in the Promised Stream ID field.
  • +
+

Receiving any frame other than HEADERS or PRIORITY on a stream in this state MUST be treated as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
reserved (local):
+
+

+
A stream in the "reserved (local)" state is one that has been promised by sending a PUSH_PROMISE frame. A PUSH_PROMISE frame reserves an idle stream by associating the stream with an open stream that was initiated by the remote peer (see Section 8.2).

+

In this state, only the following transitions are possible:

+
    +
  • The endpoint can send a HEADERS frame. This causes the stream to open in a "half closed (remote)" state.
  • +
  • Either endpoint can send a RST_STREAM frame to cause the stream to become "closed". This releases the stream reservation.
  • +
+

An endpoint MUST NOT send any type of frame other than HEADERS, RST_STREAM, or PRIORITY in this state.

+

A PRIORITY or WINDOW_UPDATE frame MAY be received in this state. Receiving any type of frame other than RST_STREAM, PRIORITY or WINDOW_UPDATE on a stream in this state MUST be treated as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
reserved (remote):
+
+

+
A stream in the "reserved (remote)" state has been reserved by a remote peer.

+

In this state, only the following transitions are possible:

+
    +
  • Receiving a HEADERS frame causes the stream to transition to "half closed (local)".
  • +
  • Either endpoint can send a RST_STREAM frame to cause the stream to become "closed". This releases the stream reservation.
  • +
+

An endpoint MAY send a PRIORITY frame in this state to reprioritize the reserved stream. An endpoint MUST NOT send any type of frame other than RST_STREAM, WINDOW_UPDATE, or PRIORITY in this state.

+

Receiving any type of frame other than HEADERS, RST_STREAM or PRIORITY on a stream in this state MUST be treated as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
open:
+
+

+
A stream in the "open" state may be used by both peers to send frames of any type. In this state, sending peers observe advertised stream level flow control limits (Section 5.2).

+

From this state either endpoint can send a frame with an END_STREAM flag set, which causes the stream to transition into one of the "half closed" states: an endpoint sending an END_STREAM flag causes the stream state to become "half closed (local)"; an endpoint receiving an END_STREAM flag causes the stream state to become "half closed (remote)".

+

Either endpoint can send a RST_STREAM frame from this state, causing it to transition immediately to "closed".

+
half closed (local):
+
+

+
A stream that is in the "half closed (local)" state cannot be used for sending frames other than WINDOW_UPDATE, PRIORITY and RST_STREAM.

+

A stream transitions from this state to "closed" when a frame that contains an END_STREAM flag is received, or when either peer sends a RST_STREAM frame.

+

An endpoint can receive any type of frame in this state. Providing flow control credit using WINDOW_UPDATE frames is necessary to continue receiving flow controlled frames. A receiver can ignore WINDOW_UPDATE frames in this state, which might arrive for a short period after a frame bearing the END_STREAM flag is sent.

+

PRIORITY frames received in this state are used to reprioritize streams that depend on the identified stream.

+
half closed (remote):
+
+

+
A stream that is "half closed (remote)" is no longer being used by the peer to send frames. In this state, an endpoint is no longer obligated to maintain a receiver flow control window.

+

If an endpoint receives additional frames for a stream that is in this state, other than WINDOW_UPDATE, PRIORITY or RST_STREAM, it MUST respond with a stream error (Section 5.4.2) of type STREAM_CLOSED.

+

A stream that is "half closed (remote)" can be used by the endpoint to send frames of any type. In this state, the endpoint continues to observe advertised stream level flow control limits (Section 5.2).

+

A stream can transition from this state to "closed" by sending a frame that contains an END_STREAM flag, or when either peer sends a RST_STREAM frame.

+
closed:
+
+

+
The "closed" state is the terminal state.

+

An endpoint MUST NOT send frames other than PRIORITY on a closed stream. An endpoint that receives any frame other than PRIORITY after receiving a RST_STREAM MUST treat that as a stream error (Section 5.4.2) of type STREAM_CLOSED. Similarly, an endpoint that receives any frames after receiving a frame with the END_STREAM flag set MUST treat that as a connection error (Section 5.4.1) of type STREAM_CLOSED, unless the frame is permitted as described below.

+

WINDOW_UPDATE or RST_STREAM frames can be received in this state for a short period after a DATA or HEADERS frame containing an END_STREAM flag is sent. Until the remote peer receives and processes RST_STREAM or the frame bearing the END_STREAM flag, it might send frames of these types. Endpoints MUST ignore WINDOW_UPDATE or RST_STREAM frames received in this state, though endpoints MAY choose to treat frames that arrive a significant time after sending END_STREAM as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+

PRIORITY frames can be sent on closed streams to prioritize streams that are dependent on the closed stream. Endpoints SHOULD process PRIORITY frames, though they can be ignored if the stream has been removed from the dependency tree (see Section 5.3.4).

+

If this state is reached as a result of sending a RST_STREAM frame, the peer that receives the RST_STREAM might have already sent - or enqueued for sending - frames on the stream that cannot be withdrawn. An endpoint MUST ignore frames that it receives on closed streams after it has sent a RST_STREAM frame. An endpoint MAY choose to limit the period over which it ignores frames and treat frames that arrive after this time as being in error.

+

Flow controlled frames (i.e., DATA) received after sending RST_STREAM are counted toward the connection flow control window. Even though these frames might be ignored, because they are sent before the sender receives the RST_STREAM, the sender will consider the frames to count against the flow control window.

+

An endpoint might receive a PUSH_PROMISE frame after it sends RST_STREAM. PUSH_PROMISE causes a stream to become "reserved" even if the associated stream has been reset. Therefore, a RST_STREAM is needed to close an unwanted promised stream.

+
+
+
+

In the absence of more specific guidance elsewhere in this document, implementations SHOULD treat the receipt of a frame that is not expressly permitted in the description of a state as a connection error (Section 5.4.1) of type PROTOCOL_ERROR. Note that PRIORITY can be sent and received in any stream state. Frames of unknown types are ignored.

+

An example of the state transitions for an HTTP request/response exchange can be found in Section 8.1. An example of the state transitions for server push can be found in Section 8.2.1 and Section 8.2.2.

+
+

5.1.1 Stream Identifiers

+

Streams are identified with an unsigned 31-bit integer. Streams initiated by a client MUST use odd-numbered stream identifiers; those initiated by the server MUST use even-numbered stream identifiers. A stream identifier of zero (0x0) is used for connection control messages; the stream identifier zero cannot be used to establish a new stream.

+

HTTP/1.1 requests that are upgraded to HTTP/2 (see Section 3.2) are responded to with a stream identifier of one (0x1). After the upgrade completes, stream 0x1 is "half closed (local)" to the client. Therefore, stream 0x1 cannot be selected as a new stream identifier by a client that upgrades from HTTP/1.1.

+

The identifier of a newly established stream MUST be numerically greater than all streams that the initiating endpoint has opened or reserved. This governs streams that are opened using a HEADERS frame and streams that are reserved using PUSH_PROMISE. An endpoint that receives an unexpected stream identifier MUST respond with a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+

The first use of a new stream identifier implicitly closes all streams in the "idle" state that might have been initiated by that peer with a lower-valued stream identifier. For example, if a client sends a HEADERS frame on stream 7 without ever sending a frame on stream 5, then stream 5 transitions to the "closed" state when the first frame for stream 7 is sent or received.

+

Stream identifiers cannot be reused. Long-lived connections can result in an endpoint exhausting the available range of stream identifiers. A client that is unable to establish a new stream identifier can establish a new connection for new streams. A server that is unable to establish a new stream identifier can send a GOAWAY frame so that the client is forced to open a new connection for new streams.

+
+
+

5.1.2 Stream Concurrency +

+

A peer can limit the number of concurrently active streams using the SETTINGS_MAX_CONCURRENT_STREAMS parameter (see Section 6.5.2) within a SETTINGS frame. The maximum concurrent streams setting is specific to each endpoint and applies only to the peer that receives the setting. That is, clients specify the maximum number of concurrent streams the server can initiate, and servers specify the maximum number of concurrent streams the client can initiate.

+

Streams that are in the "open" state, or either of the "half closed" states count toward the maximum number of streams that an endpoint is permitted to open. Streams in any of these three states count toward the limit advertised in the SETTINGS_MAX_CONCURRENT_STREAMS setting. Streams in either of the "reserved" states do not count toward the stream limit.

+

Endpoints MUST NOT exceed the limit set by their peer. An endpoint that receives a HEADERS frame that causes their advertised concurrent stream limit to be exceeded MUST treat this as a stream error (Section 5.4.2) of type PROTOCOL_ERROR or REFUSED_STREAM. The choice of error code determines whether the endpoint wishes to enable automatic retry, see Section 8.1.4) for details.

+

An endpoint that wishes to reduce the value of SETTINGS_MAX_CONCURRENT_STREAMS to a value that is below the current number of open streams can either close streams that exceed the new value or allow streams to complete.

+
+
+
+

5.2 Flow Control

+

Using streams for multiplexing introduces contention over use of the TCP connection, resulting in blocked streams. A flow control scheme ensures that streams on the same connection do not destructively interfere with each other. Flow control is used for both individual streams and for the connection as a whole.

+

HTTP/2 provides for flow control through use of the WINDOW_UPDATE frame (Section 6.9).

+
+

5.2.1 Flow Control Principles

+
+

HTTP/2 stream flow control aims to allow a variety of flow control algorithms to be used without requiring protocol changes. Flow control in HTTP/2 has the following characteristics:

+
    +
  1. Flow control is specific to a connection. Both types of flow control are between the endpoints of a single hop, and not over the entire end-to-end path.
  2. +
  3. Flow control is based on window update frames. Receivers advertise how many octets they are prepared to receive on a stream and for the entire connection. This is a credit-based scheme.
  4. +
  5. Flow control is directional with overall control provided by the receiver. A receiver MAY choose to set any window size that it desires for each stream and for the entire connection. A sender MUST respect flow control limits imposed by a receiver. Clients, servers and intermediaries all independently advertise their flow control window as a receiver and abide by the flow control limits set by their peer when sending.
  6. +
  7. The initial value for the flow control window is 65,535 octets for both new streams and the overall connection.
  8. +
  9. The frame type determines whether flow control applies to a frame. Of the frames specified in this document, only DATA frames are subject to flow control; all other frame types do not consume space in the advertised flow control window. This ensures that important control frames are not blocked by flow control.
  10. +
  11. Flow control cannot be disabled.
  12. +
  13. HTTP/2 defines only the format and semantics of the WINDOW_UPDATE frame (Section 6.9). This document does not stipulate how a receiver decides when to send this frame or the value that it sends, nor does it specify how a sender chooses to send packets. Implementations are able to select any algorithm that suits their needs.
  14. +
+
+

Implementations are also responsible for managing how requests and responses are sent based on priority; choosing how to avoid head of line blocking for requests; and managing the creation of new streams. Algorithm choices for these could interact with any flow control algorithm.

+
+
+

5.2.2 Appropriate Use of Flow Control

+

Flow control is defined to protect endpoints that are operating under resource constraints. For example, a proxy needs to share memory between many connections, and also might have a slow upstream connection and a fast downstream one. Flow control addresses cases where the receiver is unable to process data on one stream, yet wants to continue to process other streams in the same connection.

+

Deployments that do not require this capability can advertise a flow control window of the maximum size (231-1), and by maintaining this window by sending a WINDOW_UPDATE frame when any data is received. This effectively disables flow control for that receiver. Conversely, a sender is always subject to the flow control window advertised by the receiver.

+

Deployments with constrained resources (for example, memory) can employ flow control to limit the amount of memory a peer can consume. Note, however, that this can lead to suboptimal use of available network resources if flow control is enabled without knowledge of the bandwidth-delay product (see + [RFC7323] ).

+

Even with full awareness of the current bandwidth-delay product, implementation of flow control can be difficult. When using flow control, the receiver MUST read from the TCP receive buffer in a timely fashion. Failure to do so could lead to a deadlock when critical frames, such as WINDOW_UPDATE, are not read and acted upon.

+
+
+
+

5.3 Stream priority

+

A client can assign a priority for a new stream by including prioritization information in the HEADERS frame (Section 6.2) that opens the stream. At any other time, the PRIORITY frame (Section 6.3) can be used to change the priority of a stream.

+

The purpose of prioritization is to allow an endpoint to express how it would prefer its peer allocate resources when managing concurrent streams. Most importantly, priority can be used to select streams for transmitting frames when there is limited capacity for sending.

+

Streams can be prioritized by marking them as dependent on the completion of other streams (Section 5.3.1). Each dependency is assigned a relative weight, a number that is used to determine the relative proportion of available resources that are assigned to streams dependent on the same stream.

+

Explicitly setting the priority for a stream is input to a prioritization process. It does not guarantee any particular processing or transmission order for the stream relative to any other stream. An endpoint cannot force a peer to process concurrent streams in a particular order using priority. Expressing priority is therefore only ever a suggestion.

+

Prioritization information can be omitted from messages. Defaults are used prior to any explicit values being provided (Section 5.3.5).

+
+

5.3.1 Stream Dependencies

+

Each stream can be given an explicit dependency on another stream. Including a dependency expresses a preference to allocate resources to the identified stream rather than to the dependent stream.

+

A stream that is not dependent on any other stream is given a stream dependency of 0x0. In other words, the non-existent stream 0 forms the root of the tree.

+

A stream that depends on another stream is a dependent stream. The stream upon which a stream is dependent is a parent stream. A dependency on a stream that is not currently in the tree - such as a stream in the "idle" state - results in that stream being given a default priority (Section 5.3.5).

+

When assigning a dependency on another stream, the stream is added as a new dependency of the parent stream. Dependent streams that share the same parent are not ordered with respect to each other. For example, if streams B and C are dependent on stream A, and if stream D is created with a dependency on stream A, this results in a dependency order of A followed by B, C, and D in any order.

+
    A                 A
+   / \      ==>      /|\
+  B   C             B D C
+
+

Figure 3: Example of Default Dependency Creation

+

An exclusive flag allows for the insertion of a new level of dependencies. The exclusive flag causes the stream to become the sole dependency of its parent stream, causing other dependencies to become dependent on the exclusive stream. In the previous example, if stream D is created with an exclusive dependency on stream A, this results in D becoming the dependency parent of B and C.

+
                      A
+    A                 |
+   / \      ==>       D
+  B   C              / \
+                    B   C
+
+

Figure 4: Example of Exclusive Dependency Creation

+

Inside the dependency tree, a dependent stream SHOULD only be allocated resources if all of the streams that it depends on (the chain of parent streams up to 0x0) are either closed, or it is not possible to make progress on them.

+

A stream cannot depend on itself. An endpoint MUST treat this as a stream error (Section 5.4.2) of type PROTOCOL_ERROR.

+
+
+

5.3.2 Dependency Weighting +

+

All dependent streams are allocated an integer weight between 1 and 256 (inclusive).

+

Streams with the same parent SHOULD be allocated resources proportionally based on their weight. Thus, if stream B depends on stream A with weight 4, and C depends on stream A with weight 12, and if no progress can be made on A, stream B ideally receives one third of the resources allocated to stream C.

+
+
+

5.3.3 Reprioritization

+

Stream priorities are changed using the PRIORITY frame. Setting a dependency causes a stream to become dependent on the identified parent stream.

+

Dependent streams move with their parent stream if the parent is reprioritized. Setting a dependency with the exclusive flag for a reprioritized stream moves all the dependencies of the new parent stream to become dependent on the reprioritized stream.

+

If a stream is made dependent on one of its own dependencies, the formerly dependent stream is first moved to be dependent on the reprioritized stream's previous parent. The moved dependency retains its weight.

+
+

For example, consider an original dependency tree where B and C depend on A, D and E depend on C, and F depends on D. If A is made dependent on D, then D takes the place of A. All other dependency relationships stay the same, except for F, which becomes dependent on A if the reprioritization is exclusive.

    ?                ?                ?                 ?
+    |               / \               |                 |
+    A              D   A              D                 D
+   / \            /   / \            / \                |
+  B   C     ==>  F   B   C   ==>    F   A       OR      A
+     / \                 |             / \             /|\
+    D   E                E            B   C           B C F
+    |                                     |             |
+    F                                     E             E
+               (intermediate)   (non-exclusive)    (exclusive)
+
+

Figure 5: Example of Dependency Reordering

+
+
+

5.3.4 Prioritization State Management

+

When a stream is removed from the dependency tree, its dependencies can be moved to become dependent on the parent of the closed stream. The weights of new dependencies are recalculated by distributing the weight of the dependency of the closed stream proportionally based on the weights of its dependencies.

+

Streams that are removed from the dependency tree cause some prioritization information to be lost. Resources are shared between streams with the same parent stream, which means that if a stream in that set closes or becomes blocked, any spare capacity allocated to a stream is distributed to the immediate neighbors of the stream. However, if the common dependency is removed from the tree, those streams share resources with streams at the next highest level.

+

For example, assume streams A and B share a parent, and streams C and D both depend on stream A. Prior to the removal of stream A, if streams A and D are unable to proceed, then stream C receives all the resources dedicated to stream A. If stream A is removed from the tree, the weight of stream A is divided between streams C and D. If stream D is still unable to proceed, this results in stream C receiving a reduced proportion of resources. For equal starting weights, C receives one third, rather than one half, of available resources.

+

It is possible for a stream to become closed while prioritization information that creates a dependency on that stream is in transit. If a stream identified in a dependency has no associated priority information, then the dependent stream is instead assigned a default priority (Section 5.3.5). This potentially creates suboptimal prioritization, since the stream could be given a priority that is different to what is intended.

+

To avoid these problems, an endpoint SHOULD retain stream prioritization state for a period after streams become closed. The longer state is retained, the lower the chance that streams are assigned incorrect or default priority values.

+

Similarly, streams that are in the "idle" state can be assigned priority or become a parent of other streams. This allows for the creation of a grouping node in the dependency tree, which enables more flexible expressions of priority. Idle streams begin with a default priority (Section 5.3.5).

+

The retention of priority information for streams that are not counted toward the limit set by SETTINGS_MAX_CONCURRENT_STREAMS could create a large state burden for an endpoint. Therefore the amount of prioritization state that is retained MAY be limited.

+

The amount of additional state an endpoint maintains for prioritization could be dependent on load; under high load, prioritization state can be discarded to limit resource commitments. In extreme cases, an endpoint could even discard prioritization state for active or reserved streams. If a limit is applied, endpoints SHOULD maintain state for at least as many streams as allowed by their setting for SETTINGS_MAX_CONCURRENT_STREAMS. Implementations SHOULD also attempt to retain state for streams that are in active use in the priority tree.

+

An endpoint receiving a PRIORITY frame that changes the priority of a closed stream SHOULD alter the dependencies of the streams that depend on it, if it has retained enough state to do so.

+
+
+

5.3.5 Default Priorities

+

All streams are initially assigned a non-exclusive dependency on stream 0x0. Pushed streams (Section 8.2) initially depend on their associated stream. In both cases, streams are assigned a default weight of 16.

+
+
+
+

5.4 Error Handling

+
+

HTTP/2 framing permits two classes of error:

+
    +
  • An error condition that renders the entire connection unusable is a connection error.
  • +
  • An error in an individual stream is a stream error.
  • +
+
+

A list of error codes is included in Section 7.

+
+

5.4.1 Connection Error Handling

+

A connection error is any error which prevents further processing of the framing layer, or which corrupts any connection state.

+

An endpoint that encounters a connection error SHOULD first send a GOAWAY frame (Section 6.8) with the stream identifier of the last stream that it successfully received from its peer. The GOAWAY frame includes an error code that indicates why the connection is terminating. After sending the GOAWAY frame for an error condition, the endpoint MUST close the TCP connection.

+

It is possible that the GOAWAY will not be reliably received by the receiving endpoint (see + [RFC7230] , Section 6.6). In the event of a connection error, GOAWAY only provides a best effort attempt to communicate with the peer about why the connection is being terminated.

+

An endpoint can end a connection at any time. In particular, an endpoint MAY choose to treat a stream error as a connection error. Endpoints SHOULD send a GOAWAY frame when ending a connection, providing that circumstances permit it.

+
+
+

5.4.2 Stream Error Handling

+

A stream error is an error related to a specific stream that does not affect processing of other streams.

+

An endpoint that detects a stream error sends a RST_STREAM frame (Section 6.4) that contains the stream identifier of the stream where the error occurred. The RST_STREAM frame includes an error code that indicates the type of error.

+

A RST_STREAM is the last frame that an endpoint can send on a stream. The peer that sends the RST_STREAM frame MUST be prepared to receive any frames that were sent or enqueued for sending by the remote peer. These frames can be ignored, except where they modify connection state (such as the state maintained for header compression (Section 4.3), or flow control).

+

Normally, an endpoint SHOULD NOT send more than one RST_STREAM frame for any stream. However, an endpoint MAY send additional RST_STREAM frames if it receives frames on a closed stream after more than a round-trip time. This behavior is permitted to deal with misbehaving implementations.

+

An endpoint MUST NOT send a RST_STREAM in response to a RST_STREAM frame, to avoid looping.

+
+
+

5.4.3 Connection Termination +

+

If the TCP connection is closed or reset while streams remain in open or half closed states, then the affected streams cannot be automatically retried (see Section 8.1.4 for details).

+
+
+
+

5.5 Extending HTTP/2

+

HTTP/2 permits extension of the protocol. Protocol extensions can be used to provide additional services or alter any aspect of the protocol, within the limitations described in this section. Extensions are effective only within the scope of a single HTTP/2 connection.

+

This applies to the protocol elements defined in this document. This does not affect the existing options for extending HTTP, such as defining new methods, status codes, or header fields.

+

Extensions are permitted to use new frame types (Section 4.1), new settings (Section 6.5.2), or new error codes (Section 7). Registries are established for managing these extension points: frame types (Section 11.2), settings (Section 11.3) and error codes (Section 11.4).

+

Implementations MUST ignore unknown or unsupported values in all extensible protocol elements. Implementations MUST discard frames that have unknown or unsupported types. This means that any of these extension points can be safely used by extensions without prior arrangement or negotiation. However, extension frames that appear in the middle of a header block (Section 4.3) are not permitted; these MUST be treated as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+

Extensions that could change the semantics of existing protocol components MUST be negotiated before being used. For example, an extension that changes the layout of the HEADERS frame cannot be used until the peer has given a positive signal that this is acceptable. In this case, it could also be necessary to coordinate when the revised layout comes into effect. Note that treating any frame other than DATA frames as flow controlled is such a change in semantics, and can only be done through negotiation.

+

This document doesn't mandate a specific method for negotiating the use of an extension, but notes that a setting (Section 6.5.2) could be used for that purpose. If both peers set a value that indicates willingness to use the extension, then the extension can be used. If a setting is used for extension negotiation, the initial value MUST be defined in such a fashion that the extension is initially disabled.

+
+
+
+

This specification defines a number of frame types, each identified by a unique 8-bit type code. Each frame type serves a distinct purpose either in the establishment and management of the connection as a whole, or of individual streams.

+

The transmission of specific frame types can alter the state of a connection. If endpoints fail to maintain a synchronized view of the connection state, successful communication within the connection will no longer be possible. Therefore, it is important that endpoints have a shared comprehension of how the state is affected by the use any given frame.

+
+

6.1 DATA

+

DATA frames (type=0x0) convey arbitrary, variable-length sequences of octets associated with a stream. One or more DATA frames are used, for instance, to carry HTTP request or response payloads.

+

DATA frames MAY also contain padding. Padding can be added to DATA frames to obscure the size of messages. Padding is a security feature; see Section 10.7.

+
 +---------------+
+ |Pad Length? (8)|
+ +---------------+-----------------------------------------------+
+ |                            Data (*)                         ...
+ +---------------------------------------------------------------+
+ |                           Padding (*)                       ...
+ +---------------------------------------------------------------+
+
+

Figure 6: DATA Frame Payload

+
+

The DATA frame contains the following fields:

+
Pad Length:
+
An 8-bit field containing the length of the frame padding in units of octets. This field is conditional and is only present if the PADDED flag is set.
Data:
+
Application data. The amount of data is the remainder of the frame payload after subtracting the length of the other fields that are present.
Padding:
+
Padding octets that contain no application semantic value. Padding octets MUST be set to zero when sending. A receiver is not obligated to verify padding, but MAY treat non-zero padding as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.
+
+
+
+

The DATA frame defines the following flags:

+
END_STREAM (0x1):
+
Bit 0 being set indicates that this frame is the last that the endpoint will send for the identified stream. Setting this flag causes the stream to enter one of the "half closed" states or the "closed" state (Section 5.1).
PADDED (0x8):
+
Bit 3 being set indicates that the Pad Length field and any padding that it describes is present.
+
+
+

DATA frames MUST be associated with a stream. If a DATA frame is received whose stream identifier field is 0x0, the recipient MUST respond with a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+

DATA frames are subject to flow control and can only be sent when a stream is in the "open" or "half closed (remote)" states. The entire DATA frame payload is included in flow control, including Pad Length and Padding fields if present. If a DATA frame is received whose stream is not in "open" or "half closed (local)" state, the recipient MUST respond with a stream error (Section 5.4.2) of type STREAM_CLOSED.

+
+

The total number of padding octets is determined by the value of the Pad Length field. If the length of the padding is the length of the frame payload or greater, the recipient MUST treat this as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
Note:
+
A frame can be increased in size by one octet by including a Pad Length field with a value of zero.
+
+
+
+
+

6.3 PRIORITY

+

The PRIORITY frame (type=0x2) specifies the sender-advised priority of a stream (Section 5.3). It can be sent at any time for any stream, including idle or closed streams.

+
 +-+-------------------------------------------------------------+
+ |E|                  Stream Dependency (31)                     |
+ +-+-------------+-----------------------------------------------+
+ |   Weight (8)  |
+ +-+-------------+
+
+

Figure 8: PRIORITY Frame Payload

+
+

The payload of a PRIORITY frame contains the following fields:

+
E:
+
A single bit flag indicates that the stream dependency is exclusive, see Section 5.3.
Stream Dependency:
+
A 31-bit stream identifier for the stream that this stream depends on, see Section 5.3.
Weight:
+
An unsigned 8-bit integer representing a priority weight for the stream, see Section 5.3. Add one to the value to obtain a weight between 1 and 256.
+
+
+

The PRIORITY frame does not define any flags.

+

The PRIORITY frame always identifies a stream. If a PRIORITY frame is received with a stream identifier of 0x0, the recipient MUST respond with a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+

The PRIORITY frame can be sent on a stream in any state, though it cannot be sent between consecutive frames that comprise a single header block (Section 4.3). Note that this frame could arrive after processing or frame sending has completed, which would cause it to have no effect on the identified stream. For a stream that is in the "half closed (remote)" or "closed" - state, this frame can only affect processing of the identified stream and its dependent streams and not frame transmission on that stream.

+

The PRIORITY frame can be sent for a stream in the "idle" or "closed" states. This allows for the reprioritization of a group of dependent streams by altering the priority of an unused or closed parent stream.

+

A PRIORITY frame with a length other than 5 octets MUST be treated as a stream error (Section 5.4.2) of type FRAME_SIZE_ERROR.

+
+
+

6.4 RST_STREAM

+

The RST_STREAM frame (type=0x3) allows for immediate termination of a stream. RST_STREAM is sent to request cancellation of a stream, or to indicate that an error condition has occurred.

+
 +---------------------------------------------------------------+
+ |                        Error Code (32)                        |
+ +---------------------------------------------------------------+
+
+

Figure 9: RST_STREAM Frame Payload

+

The RST_STREAM frame contains a single unsigned, 32-bit integer identifying the error code (Section 7). The error code indicates why the stream is being terminated.

+

The RST_STREAM frame does not define any flags.

+

The RST_STREAM frame fully terminates the referenced stream and causes it to enter the closed state. After receiving a RST_STREAM on a stream, the receiver MUST NOT send additional frames for that stream, with the exception of PRIORITY. However, after sending the RST_STREAM, the sending endpoint MUST be prepared to receive and process additional frames sent on the stream that might have been sent by the peer prior to the arrival of the RST_STREAM.

+

RST_STREAM frames MUST be associated with a stream. If a RST_STREAM frame is received with a stream identifier of 0x0, the recipient MUST treat this as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+

RST_STREAM frames MUST NOT be sent for a stream in the "idle" state. If a RST_STREAM frame identifying an idle stream is received, the recipient MUST treat this as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+

A RST_STREAM frame with a length other than 4 octets MUST be treated as a connection error (Section 5.4.1) of type FRAME_SIZE_ERROR.

+
+
+

6.5 SETTINGS

+

The SETTINGS frame (type=0x4) conveys configuration parameters that affect how endpoints communicate, such as preferences and constraints on peer behavior. The SETTINGS frame is also used to acknowledge the receipt of those parameters. Individually, a SETTINGS parameter can also be referred to as a "setting".

+

SETTINGS parameters are not negotiated; they describe characteristics of the sending peer, which are used by the receiving peer. Different values for the same parameter can be advertised by each peer. For example, a client might set a high initial flow control window, whereas a server might set a lower value to conserve resources.

+

A SETTINGS frame MUST be sent by both endpoints at the start of a connection, and MAY be sent at any other time by either endpoint over the lifetime of the connection. Implementations MUST support all of the parameters defined by this specification.

+

Each parameter in a SETTINGS frame replaces any existing value for that parameter. Parameters are processed in the order in which they appear, and a receiver of a SETTINGS frame does not need to maintain any state other than the current value of its parameters. Therefore, the value of a SETTINGS parameter is the last value that is seen by a receiver.

+
+

SETTINGS parameters are acknowledged by the receiving peer. To enable this, the SETTINGS frame defines the following flag:

+
ACK (0x1):
+
Bit 0 being set indicates that this frame acknowledges receipt and application of the peer's SETTINGS frame. When this bit is set, the payload of the SETTINGS frame MUST be empty. Receipt of a SETTINGS frame with the ACK flag set and a length field value other than 0 MUST be treated as a connection error (Section 5.4.1) of type FRAME_SIZE_ERROR. For more info, see Settings Synchronization (Section 6.5.3).
+
+
+

SETTINGS frames always apply to a connection, never a single stream. The stream identifier for a SETTINGS frame MUST be zero (0x0). If an endpoint receives a SETTINGS frame whose stream identifier field is anything other than 0x0, the endpoint MUST respond with a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+

The SETTINGS frame affects connection state. A badly formed or incomplete SETTINGS frame MUST be treated as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+

A SETTINGS frame with a length other than a multiple of 6 octets MUST be treated as a connection error (Section 5.4.1) of type FRAME_SIZE_ERROR.

+
+

6.5.1 SETTINGS Format

+

The payload of a SETTINGS frame consists of zero or more parameters, each consisting of an unsigned 16-bit setting identifier and an unsigned 32-bit value.

+
 +-------------------------------+
+ |       Identifier (16)         |
+ +-------------------------------+-------------------------------+
+ |                        Value (32)                             |
+ +---------------------------------------------------------------+
+
+

Figure 10: Setting Format

+
+
+

6.5.2 Defined SETTINGS Parameters

+
+

The following parameters are defined:

+
+
+

Allows the sender to inform the remote endpoint of the maximum size of the header compression table used to decode header blocks, in octets. The encoder can select any size equal to or less than this value by using signaling specific to the header compression format inside a header block, see + [COMPRESSION] . The initial value is 4,096 octets.

+
SETTINGS_ENABLE_PUSH (0x2):
+
+

This setting can be use to disable server push (Section 8.2). An endpoint MUST NOT send a PUSH_PROMISE frame if it receives this parameter set to a value of 0. An endpoint that has both set this parameter to 0 and had it acknowledged MUST treat the receipt of a PUSH_PROMISE frame as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+

The initial value is 1, which indicates that server push is permitted. Any value other than 0 or 1 MUST be treated as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
SETTINGS_MAX_CONCURRENT_STREAMS (0x3):
+
+

Indicates the maximum number of concurrent streams that the sender will allow. This limit is directional: it applies to the number of streams that the sender permits the receiver to create. Initially there is no limit to this value. It is recommended that this value be no smaller than 100, so as to not unnecessarily limit parallelism.

+

A value of 0 for SETTINGS_MAX_CONCURRENT_STREAMS SHOULD NOT be treated as special by endpoints. A zero value does prevent the creation of new streams, however this can also happen for any limit that is exhausted with active streams. Servers SHOULD only set a zero value for short durations; if a server does not wish to accept requests, closing the connection is more appropriate.

+
SETTINGS_INITIAL_WINDOW_SIZE (0x4):
+
+

Indicates the sender's initial window size (in octets) for stream level flow control. The initial value is 216-1 (65,535) octets.

+

This setting affects the window size of all streams, see Section 6.9.2.

+

Values above the maximum flow control window size of 231-1 MUST be treated as a connection error (Section 5.4.1) of type FLOW_CONTROL_ERROR.

+
SETTINGS_MAX_FRAME_SIZE (0x5):
+
+

Indicates the size of the largest frame payload that the sender is willing to receive, in octets.

+

The initial value is 214 (16,384) octets. The value advertised by an endpoint MUST be between this initial value and the maximum allowed frame size (224-1 or 16,777,215 octets), inclusive. Values outside this range MUST be treated as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
+
+

This advisory setting informs a peer of the maximum size of header list that the sender is prepared to accept, in octets. The value is based on the uncompressed size of header fields, including the length of the name and value in octets plus an overhead of 32 octets for each header field.

+

For any given request, a lower limit than what is advertised MAY be enforced. The initial value of this setting is unlimited.

+
+
+
+

An endpoint that receives a SETTINGS frame with any unknown or unsupported identifier MUST ignore that setting.

+
+
+

6.5.3 Settings Synchronization

+

Most values in SETTINGS benefit from or require an understanding of when the peer has received and applied the changed parameter values. In order to provide such synchronization timepoints, the recipient of a SETTINGS frame in which the ACK flag is not set MUST apply the updated parameters as soon as possible upon receipt.

+

The values in the SETTINGS frame MUST be processed in the order they appear, with no other frame processing between values. Unsupported parameters MUST be ignored. Once all values have been processed, the recipient MUST immediately emit a SETTINGS frame with the ACK flag set. Upon receiving a SETTINGS frame with the ACK flag set, the sender of the altered parameters can rely on the setting having been applied.

+

If the sender of a SETTINGS frame does not receive an acknowledgement within a reasonable amount of time, it MAY issue a connection error (Section 5.4.1) of type SETTINGS_TIMEOUT.

+
+
+
+

6.6 PUSH_PROMISE

+

The PUSH_PROMISE frame (type=0x5) is used to notify the peer endpoint in advance of streams the sender intends to initiate. The PUSH_PROMISE frame includes the unsigned 31-bit identifier of the stream the endpoint plans to create along with a set of headers that provide additional context for the stream. Section 8.2 contains a thorough description of the use of PUSH_PROMISE frames.

+
 +---------------+
+ |Pad Length? (8)|
+ +-+-------------+-----------------------------------------------+
+ |R|                  Promised Stream ID (31)                    |
+ +-+-----------------------------+-------------------------------+
+ |                   Header Block Fragment (*)                 ...
+ +---------------------------------------------------------------+
+ |                           Padding (*)                       ...
+ +---------------------------------------------------------------+
+
+

Figure 11: PUSH_PROMISE Payload Format

+
+

The PUSH_PROMISE frame payload has the following fields:

+
Pad Length:
+
An 8-bit field containing the length of the frame padding in units of octets. This field is only present if the PADDED flag is set.
R:
+
A single reserved bit.
Promised Stream ID:
+
An unsigned 31-bit integer that identifies the stream that is reserved by the PUSH_PROMISE. The promised stream identifier MUST be a valid choice for the next stream sent by the sender (see new stream identifier (Section 5.1.1)).
Header Block Fragment:
+
A header block fragment (Section 4.3) containing request header fields.
Padding:
+
Padding octets.
+
+
+
+

The PUSH_PROMISE frame defines the following flags:

+
END_HEADERS (0x4):
+
+

Bit 2 being set indicates that this frame contains an entire header block (Section 4.3) and is not followed by any CONTINUATION frames.

+

A PUSH_PROMISE frame without the END_HEADERS flag set MUST be followed by a CONTINUATION frame for the same stream. A receiver MUST treat the receipt of any other type of frame or a frame on a different stream as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
PADDED (0x8):
+
+

Bit 3 being set indicates that the Pad Length field and any padding that it describes is present.

+
+
+
+

PUSH_PROMISE frames MUST only be sent on a peer-initiated stream that is in either the "open" or "half closed (remote)" state. The stream identifier of a PUSH_PROMISE frame indicates the stream it is associated with. If the stream identifier field specifies the value 0x0, a recipient MUST respond with a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+

Promised streams are not required to be used in the order they are promised. The PUSH_PROMISE only reserves stream identifiers for later use.

+

PUSH_PROMISE MUST NOT be sent if the SETTINGS_ENABLE_PUSH setting of the peer endpoint is set to 0. An endpoint that has set this setting and has received acknowledgement MUST treat the receipt of a PUSH_PROMISE frame as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+

Recipients of PUSH_PROMISE frames can choose to reject promised streams by returning a RST_STREAM referencing the promised stream identifier back to the sender of the PUSH_PROMISE.

+

A PUSH_PROMISE frame modifies the connection state in two ways. The inclusion of a header block (Section 4.3) potentially modifies the state maintained for header compression. PUSH_PROMISE also reserves a stream for later use, causing the promised stream to enter the "reserved" state. A sender MUST NOT send a PUSH_PROMISE on a stream unless that stream is either "open" or "half closed (remote)"; the sender MUST ensure that the promised stream is a valid choice for a new stream identifier (Section 5.1.1) (that is, the promised stream MUST be in the "idle" state).

+

Since PUSH_PROMISE reserves a stream, ignoring a PUSH_PROMISE frame causes the stream state to become indeterminate. A receiver MUST treat the receipt of a PUSH_PROMISE on a stream that is neither "open" nor "half closed (local)" as a connection error (Section 5.4.1) of type PROTOCOL_ERROR. However, an endpoint that has sent RST_STREAM on the associated stream MUST handle PUSH_PROMISE frames that might have been created before the RST_STREAM frame is received and processed.

+

A receiver MUST treat the receipt of a PUSH_PROMISE that promises an illegal stream identifier (Section 5.1.1) (that is, an identifier for a stream that is not currently in the "idle" state) as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+

The PUSH_PROMISE frame can include padding. Padding fields and flags are identical to those defined for DATA frames (Section 6.1).

+
+
+

6.7 PING

+

The PING frame (type=0x6) is a mechanism for measuring a minimal round trip time from the sender, as well as determining whether an idle connection is still functional. PING frames can be sent from any endpoint.

+
 +---------------------------------------------------------------+
+ |                                                               |
+ |                      Opaque Data (64)                         |
+ |                                                               |
+ +---------------------------------------------------------------+
+
+

Figure 12: PING Payload Format

+

In addition to the frame header, PING frames MUST contain 8 octets of data in the payload. A sender can include any value it chooses and use those octets in any fashion.

+

Receivers of a PING frame that does not include an ACK flag MUST send a PING frame with the ACK flag set in response, with an identical payload. PING responses SHOULD be given higher priority than any other frame.

+
+

The PING frame defines the following flags:

+
ACK (0x1):
+
Bit 0 being set indicates that this PING frame is a PING response. An endpoint MUST set this flag in PING responses. An endpoint MUST NOT respond to PING frames containing this flag.
+
+
+

PING frames are not associated with any individual stream. If a PING frame is received with a stream identifier field value other than 0x0, the recipient MUST respond with a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+

Receipt of a PING frame with a length field value other than 8 MUST be treated as a connection error (Section 5.4.1) of type FRAME_SIZE_ERROR.

+
+
+

6.8 GOAWAY

+

The GOAWAY frame (type=0x7) informs the remote peer to stop creating streams on this connection. GOAWAY can be sent by either the client or the server. Once sent, the sender will ignore frames sent on any new streams with identifiers higher than the included last stream identifier. Receivers of a GOAWAY frame MUST NOT open additional streams on the connection, although a new connection can be established for new streams.

+

The purpose of this frame is to allow an endpoint to gracefully stop accepting new streams, while still finishing processing of previously established streams. This enables administrative actions, like server maintenance.

+

There is an inherent race condition between an endpoint starting new streams and the remote sending a GOAWAY frame. To deal with this case, the GOAWAY contains the stream identifier of the last peer-initiated stream which was or might be processed on the sending endpoint in this connection. For instance, if the server sends a GOAWAY frame, the identified stream is the highest numbered stream initiated by the client.

+

If the receiver of the GOAWAY has sent data on streams with a higher stream identifier than what is indicated in the GOAWAY frame, those streams are not or will not be processed. The receiver of the GOAWAY frame can treat the streams as though they had never been created at all, thereby allowing those streams to be retried later on a new connection.

+

Endpoints SHOULD always send a GOAWAY frame before closing a connection so that the remote peer can know whether a stream has been partially processed or not. For example, if an HTTP client sends a POST at the same time that a server closes a connection, the client cannot know if the server started to process that POST request if the server does not send a GOAWAY frame to indicate what streams it might have acted on.

+

An endpoint might choose to close a connection without sending GOAWAY for misbehaving peers.

+
 +-+-------------------------------------------------------------+
+ |R|                  Last-Stream-ID (31)                        |
+ +-+-------------------------------------------------------------+
+ |                      Error Code (32)                          |
+ +---------------------------------------------------------------+
+ |                  Additional Debug Data (*)                    |
+ +---------------------------------------------------------------+
+
+

Figure 13: GOAWAY Payload Format

+

The GOAWAY frame does not define any flags.

+

The GOAWAY frame applies to the connection, not a specific stream. An endpoint MUST treat a GOAWAY frame with a stream identifier other than 0x0 as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
+

The last stream identifier in the GOAWAY frame contains the highest numbered stream identifier for which the sender of the GOAWAY frame might have taken some action on, or might yet take action on. All streams up to and including the identified stream might have been processed in some way. The last stream identifier can be set to 0 if no streams were processed.

+
Note:
+
In this context, "processed" means that some data from the stream was passed to some higher layer of software that might have taken some action as a result.
+
+

If a connection terminates without a GOAWAY frame, the last stream identifier is effectively the highest possible stream identifier.

+
+

On streams with lower or equal numbered identifiers that were not closed completely prior to the connection being closed, re-attempting requests, transactions, or any protocol activity is not possible, with the exception of idempotent actions like HTTP GET, PUT, or DELETE. Any protocol activity that uses higher numbered streams can be safely retried using a new connection.

+

Activity on streams numbered lower or equal to the last stream identifier might still complete successfully. The sender of a GOAWAY frame might gracefully shut down a connection by sending a GOAWAY frame, maintaining the connection in an open state until all in-progress streams complete.

+

An endpoint MAY send multiple GOAWAY frames if circumstances change. For instance, an endpoint that sends GOAWAY with NO_ERROR during graceful shutdown could subsequently encounter a condition that requires immediate termination of the connection. The last stream identifier from the last GOAWAY frame received indicates which streams could have been acted upon. Endpoints MUST NOT increase the value they send in the last stream identifier, since the peers might already have retried unprocessed requests on another connection.

+

A client that is unable to retry requests loses all requests that are in flight when the server closes the connection. This is especially true for intermediaries that might not be serving clients using HTTP/2. A server that is attempting to gracefully shut down a connection SHOULD send an initial GOAWAY frame with the last stream identifier set to 231-1 and a NO_ERROR code. This signals to the client that a shutdown is imminent and that no further requests can be initiated. After waiting at least one round trip time, the server can send another GOAWAY frame with an updated last stream identifier. This ensures that a connection can be cleanly shut down without losing requests.

+

After sending a GOAWAY frame, the sender can discard frames for streams with identifiers higher than the identified last stream. However, any frames that alter connection state cannot be completely ignored. For instance, HEADERS, PUSH_PROMISE and CONTINUATION frames MUST be minimally processed to ensure the state maintained for header compression is consistent (see Section 4.3); similarly DATA frames MUST be counted toward the connection flow control window. Failure to process these frames can cause flow control or header compression state to become unsynchronized.

+

The GOAWAY frame also contains a 32-bit error code (Section 7) that contains the reason for closing the connection.

+

Endpoints MAY append opaque data to the payload of any GOAWAY frame. Additional debug data is intended for diagnostic purposes only and carries no semantic value. Debug information could contain security- or privacy-sensitive data. Logged or otherwise persistently stored debug data MUST have adequate safeguards to prevent unauthorized access.

+
+
+

6.9 WINDOW_UPDATE

+

The WINDOW_UPDATE frame (type=0x8) is used to implement flow control; see Section 5.2 for an overview.

+

Flow control operates at two levels: on each individual stream and on the entire connection.

+

Both types of flow control are hop-by-hop; that is, only between the two endpoints. Intermediaries do not forward WINDOW_UPDATE frames between dependent connections. However, throttling of data transfer by any receiver can indirectly cause the propagation of flow control information toward the original sender.

+

Flow control only applies to frames that are identified as being subject to flow control. Of the frame types defined in this document, this includes only DATA frames. Frames that are exempt from flow control MUST be accepted and processed, unless the receiver is unable to assign resources to handling the frame. A receiver MAY respond with a stream error (Section 5.4.2) or connection error (Section 5.4.1) of type FLOW_CONTROL_ERROR if it is unable to accept a frame.

+
 +-+-------------------------------------------------------------+
+ |R|              Window Size Increment (31)                     |
+ +-+-------------------------------------------------------------+
+
+

Figure 14: WINDOW_UPDATE Payload Format

+

The payload of a WINDOW_UPDATE frame is one reserved bit, plus an unsigned 31-bit integer indicating the number of octets that the sender can transmit in addition to the existing flow control window. The legal range for the increment to the flow control window is 1 to 231-1 (2,147,483,647) octets.

+

The WINDOW_UPDATE frame does not define any flags.

+

The WINDOW_UPDATE frame can be specific to a stream or to the entire connection. In the former case, the frame's stream identifier indicates the affected stream; in the latter, the value "0" indicates that the entire connection is the subject of the frame.

+

A receiver MUST treat the receipt of a WINDOW_UPDATE frame with an flow control window increment of 0 as a stream error (Section 5.4.2) of type PROTOCOL_ERROR; errors on the connection flow control window MUST be treated as a connection error (Section 5.4.1).

+

WINDOW_UPDATE can be sent by a peer that has sent a frame bearing the END_STREAM flag. This means that a receiver could receive a WINDOW_UPDATE frame on a "half closed (remote)" or "closed" stream. A receiver MUST NOT treat this as an error, see Section 5.1.

+

A receiver that receives a flow controlled frame MUST always account for its contribution against the connection flow control window, unless the receiver treats this as a connection error (Section 5.4.1). This is necessary even if the frame is in error. Since the sender counts the frame toward the flow control window, if the receiver does not, the flow control window at sender and receiver can become different.

+

A WINDOW_UPDATE frame with a length other than 4 octets MUST be treated as a connection error (Section 5.4.1) of type FRAME_SIZE_ERROR.

+
+

6.9.1 The Flow Control Window +

+

Flow control in HTTP/2 is implemented using a window kept by each sender on every stream. The flow control window is a simple integer value that indicates how many octets of data the sender is permitted to transmit; as such, its size is a measure of the buffering capacity of the receiver.

+

Two flow control windows are applicable: the stream flow control window and the connection flow control window. The sender MUST NOT send a flow controlled frame with a length that exceeds the space available in either of the flow control windows advertised by the receiver. Frames with zero length with the END_STREAM flag set (that is, an empty DATA frame) MAY be sent if there is no available space in either flow control window.

+

For flow control calculations, the 9 octet frame header is not counted.

+

After sending a flow controlled frame, the sender reduces the space available in both windows by the length of the transmitted frame.

+

The receiver of a frame sends a WINDOW_UPDATE frame as it consumes data and frees up space in flow control windows. Separate WINDOW_UPDATE frames are sent for the stream and connection level flow control windows.

+

A sender that receives a WINDOW_UPDATE frame updates the corresponding window by the amount specified in the frame.

+

A sender MUST NOT allow a flow control window to exceed 231-1 octets. If a sender receives a WINDOW_UPDATE that causes a flow control window to exceed this maximum it MUST terminate either the stream or the connection, as appropriate. For streams, the sender sends a RST_STREAM with the error code of FLOW_CONTROL_ERROR code; for the connection, a GOAWAY frame with a FLOW_CONTROL_ERROR code.

+

Flow controlled frames from the sender and WINDOW_UPDATE frames from the receiver are completely asynchronous with respect to each other. This property allows a receiver to aggressively update the window size kept by the sender to prevent streams from stalling.

+
+
+

6.9.2 Initial Flow Control Window Size

+

When an HTTP/2 connection is first established, new streams are created with an initial flow control window size of 65,535 octets. The connection flow control window is 65,535 octets. Both endpoints can adjust the initial window size for new streams by including a value for SETTINGS_INITIAL_WINDOW_SIZE in the SETTINGS frame that forms part of the connection preface. The connection flow control window can only be changed using WINDOW_UPDATE frames.

+

Prior to receiving a SETTINGS frame that sets a value for SETTINGS_INITIAL_WINDOW_SIZE, an endpoint can only use the default initial window size when sending flow controlled frames. Similarly, the connection flow control window is set to the default initial window size until a WINDOW_UPDATE frame is received.

+

A SETTINGS frame can alter the initial flow control window size for all streams in the "open" or "half closed (remote)" state. When the value of SETTINGS_INITIAL_WINDOW_SIZE changes, a receiver MUST adjust the size of all stream flow control windows that it maintains by the difference between the new value and the old value.

+

A change to SETTINGS_INITIAL_WINDOW_SIZE can cause the available space in a flow control window to become negative. A sender MUST track the negative flow control window, and MUST NOT send new flow controlled frames until it receives WINDOW_UPDATE frames that cause the flow control window to become positive.

+

For example, if the client sends 60KB immediately on connection establishment, and the server sets the initial window size to be 16KB, the client will recalculate the available flow control window to be -44KB on receipt of the SETTINGS frame. The client retains a negative flow control window until WINDOW_UPDATE frames restore the window to being positive, after which the client can resume sending.

+

A SETTINGS frame cannot alter the connection flow control window.

+

An endpoint MUST treat a change to SETTINGS_INITIAL_WINDOW_SIZE that causes any flow control window to exceed the maximum size as a connection error (Section 5.4.1) of type FLOW_CONTROL_ERROR.

+
+
+

6.9.3 Reducing the Stream Window Size +

+

A receiver that wishes to use a smaller flow control window than the current size can send a new SETTINGS frame. However, the receiver MUST be prepared to receive data that exceeds this window size, since the sender might send data that exceeds the lower limit prior to processing the SETTINGS frame.

+

After sending a SETTINGS frame that reduces the initial flow control window size, a receiver MAY continue to process streams that exceed flow control limits. Allowing streams to continue does not allow the receiver to immediately reduce the space it reserves for flow control windows. Progress on these streams can also stall, since WINDOW_UPDATE frames are needed to allow the sender to resume sending. The receiver MAY instead send a RST_STREAM with FLOW_CONTROL_ERROR error code for the affected streams.

+
+
+
+

6.10 CONTINUATION

+

The CONTINUATION frame (type=0x9) is used to continue a sequence of header block fragments (Section 4.3). Any number of CONTINUATION frames can be sent, as long as the preceding frame is on the same stream and is a HEADERS, PUSH_PROMISE or CONTINUATION frame without the END_HEADERS flag set.

+
 +---------------------------------------------------------------+
+ |                   Header Block Fragment (*)                 ...
+ +---------------------------------------------------------------+
+
+

Figure 15: CONTINUATION Frame Payload

+

The CONTINUATION frame payload contains a header block fragment (Section 4.3).

+
+

The CONTINUATION frame defines the following flag:

+
END_HEADERS (0x4):
+
+

Bit 2 being set indicates that this frame ends a header block (Section 4.3).

+

If the END_HEADERS bit is not set, this frame MUST be followed by another CONTINUATION frame. A receiver MUST treat the receipt of any other type of frame or a frame on a different stream as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
+
+
+

The CONTINUATION frame changes the connection state as defined in Section 4.3.

+

CONTINUATION frames MUST be associated with a stream. If a CONTINUATION frame is received whose stream identifier field is 0x0, the recipient MUST respond with a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+

A CONTINUATION frame MUST be preceded by a HEADERS, PUSH_PROMISE or CONTINUATION frame without the END_HEADERS flag set. A recipient that observes violation of this rule MUST respond with a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
+
+
+

Error codes are 32-bit fields that are used in RST_STREAM and GOAWAY frames to convey the reasons for the stream or connection error.

+

Error codes share a common code space. Some error codes apply only to either streams or the entire connection and have no defined semantics in the other context.

+
+

The following error codes are defined:

+
NO_ERROR (0x0):
+
The associated condition is not as a result of an error. For example, a GOAWAY might include this code to indicate graceful shutdown of a connection.
PROTOCOL_ERROR (0x1):
+
The endpoint detected an unspecific protocol error. This error is for use when a more specific error code is not available.
INTERNAL_ERROR (0x2):
+
The endpoint encountered an unexpected internal error.
FLOW_CONTROL_ERROR (0x3):
+
The endpoint detected that its peer violated the flow control protocol.
SETTINGS_TIMEOUT (0x4):
+
The endpoint sent a SETTINGS frame, but did not receive a response in a timely manner. See Settings Synchronization (Section 6.5.3).
STREAM_CLOSED (0x5):
+
The endpoint received a frame after a stream was half closed.
FRAME_SIZE_ERROR (0x6):
+
The endpoint received a frame with an invalid size.
REFUSED_STREAM (0x7):
+
The endpoint refuses the stream prior to performing any application processing, see Section 8.1.4 for details.
CANCEL (0x8):
+
Used by the endpoint to indicate that the stream is no longer needed.
COMPRESSION_ERROR (0x9):
+
The endpoint is unable to maintain the header compression context for the connection.
CONNECT_ERROR (0xa):
+
The connection established in response to a CONNECT request (Section 8.3) was reset or abnormally closed.
ENHANCE_YOUR_CALM (0xb):
+
The endpoint detected that its peer is exhibiting a behavior that might be generating excessive load.
INADEQUATE_SECURITY (0xc):
+
The underlying transport has properties that do not meet minimum security requirements (see Section 9.2).
HTTP_1_1_REQUIRED (0xd):
+
The endpoint requires that HTTP/1.1 be used instead of HTTP/2.
+
+
+

Unknown or unsupported error codes MUST NOT trigger any special behavior. These MAY be treated by an implementation as being equivalent to INTERNAL_ERROR.

+
+
+

HTTP/2 is intended to be as compatible as possible with current uses of HTTP. This means that, from the application perspective, the features of the protocol are largely unchanged. To achieve this, all request and response semantics are preserved, although the syntax of conveying those semantics has changed.

+

Thus, the specification and requirements of HTTP/1.1 Semantics and Content + [RFC7231] , Conditional Requests + [RFC7232] , Range Requests + [RFC7233] , Caching + [RFC7234] and Authentication + [RFC7235] are applicable to HTTP/2. Selected portions of HTTP/1.1 Message Syntax and Routing + [RFC7230] , such as the HTTP and HTTPS URI schemes, are also applicable in HTTP/2, but the expression of those semantics for this protocol are defined in the sections below.

+
+

8.1 HTTP Request/Response Exchange

+

A client sends an HTTP request on a new stream, using a previously unused stream identifier (Section 5.1.1). A server sends an HTTP response on the same stream as the request.

+
+

An HTTP message (request or response) consists of:

+
    +
  1. for a response only, zero or more HEADERS frames (each followed by zero or more CONTINUATION frames) containing the message headers of informational (1xx) HTTP responses (see + [RFC7230] , Section 3.2 and + [RFC7231] , Section 6.2), and
  2. +
  3. one HEADERS frame (followed by zero or more CONTINUATION frames) containing the message headers (see + [RFC7230] , Section 3.2), and
  4. +
  5. zero or more DATA frames containing the payload body (see + [RFC7230] , Section 3.3), and
  6. +
  7. optionally, one HEADERS frame, followed by zero or more CONTINUATION frames containing the trailer-part, if present (see + [RFC7230] , Section 4.1.2).
  8. +
+

The last frame in the sequence bears an END_STREAM flag, noting that a HEADERS frame bearing the END_STREAM flag can be followed by CONTINUATION frames that carry any remaining portions of the header block.

+
+

Other frames (from any stream) MUST NOT occur between either HEADERS frame and any CONTINUATION frames that might follow.

+

HTTP/2 uses DATA frames to carry message payloads. The chunked transfer encoding defined in Section 4.1 of + [RFC7230] MUST NOT be used in HTTP/2.

+

Trailing header fields are carried in a header block that also terminates the stream. Such a header block is a sequence starting with a HEADERS frame, followed by zero or more CONTINUATION frames, where the HEADERS frame bears an END_STREAM flag. Header blocks after the first that do not terminate the stream are not part of an HTTP request or response.

+

A HEADERS frame (and associated CONTINUATION frames) can only appear at the start or end of a stream. An endpoint that receives a HEADERS frame without the END_STREAM flag set after receiving a final (non-informational) status code MUST treat the corresponding request or response as malformed (Section 8.1.2.6).

+

An HTTP request/response exchange fully consumes a single stream. A request starts with the HEADERS frame that puts the stream into an "open" state. The request ends with a frame bearing END_STREAM, which causes the stream to become "half closed (local)" for the client and "half closed (remote)" for the server. A response starts with a HEADERS frame and ends with a frame bearing END_STREAM, which places the stream in the "closed" state.

+

An HTTP response is complete after the server sends - or the client receives - a frame with the END_STREAM flag set (including any CONTINUATION frames needed to complete a header block). A server can send a complete response prior to the client sending an entire request if the response does not depend on any portion of the request that has not been sent and received. When this is true, a server MAY request that the client abort transmission of a request without error by sending a RST_STREAM with an error code of NO_ERROR after sending a complete response (i.e., a frame with the END_STREAM flag). Clients MUST NOT discard responses as a result of receiving such a RST_STREAM, though clients can always discard responses at their discretion for other reasons.

+
+

8.1.1 Upgrading From HTTP/2

+

HTTP/2 removes support for the 101 (Switching Protocols) informational status code ( + [RFC7231] , Section 6.2.2).

+

The semantics of 101 (Switching Protocols) aren't applicable to a multiplexed protocol. Alternative protocols are able to use the same mechanisms that HTTP/2 uses to negotiate their use (see Section 3).

+
+
+

8.1.3 Examples +

+

This section shows HTTP/1.1 requests and responses, with illustrations of equivalent HTTP/2 requests and responses.

+

An HTTP GET request includes request header fields and no payload body and is therefore transmitted as a single HEADERS frame, followed by zero or more CONTINUATION frames containing the serialized block of request header fields. The HEADERS frame in the following has both the END_HEADERS and END_STREAM flags set; no CONTINUATION frames are sent:

+
  GET /resource HTTP/1.1           HEADERS
+  Host: example.org          ==>     + END_STREAM
+  Accept: image/jpeg                 + END_HEADERS
+                                       :method = GET
+                                       :scheme = https
+                                       :path = /resource
+                                       host = example.org
+                                       accept = image/jpeg
+
+

Similarly, a response that includes only response header fields is transmitted as a HEADERS frame (again, followed by zero or more CONTINUATION frames) containing the serialized block of response header fields.

+
  HTTP/1.1 304 Not Modified        HEADERS
+  ETag: "xyzzy"              ==>     + END_STREAM
+  Expires: Thu, 23 Jan ...           + END_HEADERS
+                                       :status = 304
+                                       etag = "xyzzy"
+                                       expires = Thu, 23 Jan ...
+
+

An HTTP POST request that includes request header fields and payload data is transmitted as one HEADERS frame, followed by zero or more CONTINUATION frames containing the request header fields, followed by one or more DATA frames, with the last CONTINUATION (or HEADERS) frame having the END_HEADERS flag set and the final DATA frame having the END_STREAM flag set:

+
  POST /resource HTTP/1.1          HEADERS
+  Host: example.org          ==>     - END_STREAM
+  Content-Type: image/jpeg           - END_HEADERS
+  Content-Length: 123                  :method = POST
+                                       :path = /resource
+  {binary data}                        :scheme = https
+
+                                   CONTINUATION
+                                     + END_HEADERS
+                                       content-type = image/jpeg
+                                       host = example.org
+                                       content-length = 123
+
+                                   DATA
+                                     + END_STREAM
+                                   {binary data}
+
+

Note that data contributing to any given header field could be spread between header block fragments. The allocation of header fields to frames in this example is illustrative only.

+
+

A response that includes header fields and payload data is transmitted as a HEADERS frame, followed by zero or more CONTINUATION frames, followed by one or more DATA frames, with the last DATA frame in the sequence having the END_STREAM flag set:

+
  HTTP/1.1 200 OK                  HEADERS
+  Content-Type: image/jpeg   ==>     - END_STREAM
+  Content-Length: 123                + END_HEADERS
+                                       :status = 200
+  {binary data}                        content-type = image/jpeg
+                                       content-length = 123
+
+                                   DATA
+                                     + END_STREAM
+                                   {binary data}
+
+

An informational response using a 1xx status code other than 101 is transmitted as a HEADERS frame, followed by zero or more CONTINUATION frames.

+

Trailing header fields are sent as a header block after both the request or response header block and all the DATA frames have been sent. The HEADERS frame starting the trailers header block has the END_STREAM flag set.

+
+

The following example includes both a 100 (Continue) status code, which is sent in response to a request containing a "100-continue" token in the Expect header field, and trailing header fields:

  HTTP/1.1 100 Continue            HEADERS
+  Extension-Field: bar       ==>     - END_STREAM
+                                     + END_HEADERS
+                                       :status = 100
+                                       extension-field = bar
+
+  HTTP/1.1 200 OK                  HEADERS
+  Content-Type: image/jpeg   ==>     - END_STREAM
+  Transfer-Encoding: chunked         + END_HEADERS
+  Trailer: Foo                         :status = 200
+                                       content-length = 123
+  123                                  content-type = image/jpeg
+  {binary data}                        trailer = Foo
+  0
+  Foo: bar                         DATA
+                                     - END_STREAM
+                                   {binary data}
+
+                                   HEADERS
+                                     + END_STREAM
+                                     + END_HEADERS
+                                       foo = bar
+
+
+
+

8.1.4 Request Reliability Mechanisms in HTTP/2

+

In HTTP/1.1, an HTTP client is unable to retry a non-idempotent request when an error occurs, because there is no means to determine the nature of the error. It is possible that some server processing occurred prior to the error, which could result in undesirable effects if the request were reattempted.

+
+

HTTP/2 provides two mechanisms for providing a guarantee to a client that a request has not been processed:

+
    +
  • The GOAWAY frame indicates the highest stream number that might have been processed. Requests on streams with higher numbers are therefore guaranteed to be safe to retry.
  • +
  • The REFUSED_STREAM error code can be included in a RST_STREAM frame to indicate that the stream is being closed prior to any processing having occurred. Any request that was sent on the reset stream can be safely retried.
  • +
+
+

Requests that have not been processed have not failed; clients MAY automatically retry them, even those with non-idempotent methods.

+

A server MUST NOT indicate that a stream has not been processed unless it can guarantee that fact. If frames that are on a stream are passed to the application layer for any stream, then REFUSED_STREAM MUST NOT be used for that stream, and a GOAWAY frame MUST include a stream identifier that is greater than or equal to the given stream identifier.

+

In addition to these mechanisms, the PING frame provides a way for a client to easily test a connection. Connections that remain idle can become broken as some middleboxes (for instance, network address translators, or load balancers) silently discard connection bindings. The PING frame allows a client to safely test whether a connection is still active without sending a request.

+
+
+
+

8.2 Server Push

+

HTTP/2 allows a server to pre-emptively send (or "push") responses (along with corresponding "promised" requests) to a client in association with a previous client-initiated request. This can be useful when the server knows the client will need to have those responses available in order to fully process the response to the original request.

+

A client can request that server push be disabled, though this is negotiated for each hop independently. The SETTINGS_ENABLE_PUSH setting can be set to 0 to indicate that server push is disabled.

+

Promised requests MUST be cacheable (see + [RFC7231] , Section 4.2.3), MUST be safe (see + [RFC7231] , Section 4.2.1) and MUST NOT include a request body. Clients that receive a promised request that is not cacheable, is not known to be safe or that indicates the presence of a request body MUST reset the promised stream with a stream error (Section 5.4.2) of type PROTOCOL_ERROR. Note this could result in the promised stream being reset if the client does not recognize a newly defined method as being safe.

+

Pushed responses that are cacheable (see + [RFC7234] , Section 3) can be stored by the client, if it implements an HTTP cache. Pushed responses are considered successfully validated on the origin server (e.g., if the "no-cache" cache response directive + [RFC7234] , Section 5.2.2 is present) while the stream identified by the promised stream ID is still open.

+

Pushed responses that are not cacheable MUST NOT be stored by any HTTP cache. They MAY be made available to the application separately.

+

The server MUST include a value in the :authority header field for which the server is authoritative (see Section 10.1). A client MUST treat a PUSH_PROMISE for which the server is not authoritative as a stream error (Section 5.4.2) of type PROTOCOL_ERROR.

+

An intermediary can receive pushes from the server and choose not to forward them on to the client. In other words, how to make use of the pushed information is up to that intermediary. Equally, the intermediary might choose to make additional pushes to the client, without any action taken by the server.

+

A client cannot push. Thus, servers MUST treat the receipt of a PUSH_PROMISE frame as a connection error (Section 5.4.1) of type PROTOCOL_ERROR. Clients MUST reject any attempt to change the SETTINGS_ENABLE_PUSH setting to a value other than 0 by treating the message as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
+

8.2.1 Push Requests

+

Server push is semantically equivalent to a server responding to a request; however, in this case that request is also sent by the server, as a PUSH_PROMISE frame.

+

The PUSH_PROMISE frame includes a header block that contains a complete set of request header fields that the server attributes to the request. It is not possible to push a response to a request that includes a request body.

+

Pushed responses are always associated with an explicit request from the client. The PUSH_PROMISE frames sent by the server are sent on that explicit request's stream. The PUSH_PROMISE frame also includes a promised stream identifier, chosen from the stream identifiers available to the server (see Section 5.1.1).

+

The header fields in PUSH_PROMISE and any subsequent CONTINUATION frames MUST be a valid and complete set of request header fields (Section 8.1.2.3). The server MUST include a method in the :method header field that is safe and cacheable. If a client receives a PUSH_PROMISE that does not include a complete and valid set of header fields, or the :method header field identifies a method that is not safe, it MUST respond with a stream error (Section 5.4.2) of type PROTOCOL_ERROR.

+

The server SHOULD send PUSH_PROMISE (Section 6.6) frames prior to sending any frames that reference the promised responses. This avoids a race where clients issue requests prior to receiving any PUSH_PROMISE frames.

+

For example, if the server receives a request for a document containing embedded links to multiple image files, and the server chooses to push those additional images to the client, sending push promises before the DATA frames that contain the image links ensures that the client is able to see the promises before discovering embedded links. Similarly, if the server pushes responses referenced by the header block (for instance, in Link header fields), sending the push promises before sending the header block ensures that clients do not request them.

+

PUSH_PROMISE frames MUST NOT be sent by the client.

+

PUSH_PROMISE frames can be sent by the server in response to any client-initiated stream, but the stream MUST be in either the "open" or "half closed (remote)" state with respect to the server. PUSH_PROMISE frames are interspersed with the frames that comprise a response, though they cannot be interspersed with HEADERS and CONTINUATION frames that comprise a single header block.

+

Sending a PUSH_PROMISE frame creates a new stream and puts the stream into the “reserved (local)” state for the server and the “reserved (remote)” state for the client.

+
+
+

8.2.2 Push Responses

+

After sending the PUSH_PROMISE frame, the server can begin delivering the pushed response as a response (Section 8.1.2.4) on a server-initiated stream that uses the promised stream identifier. The server uses this stream to transmit an HTTP response, using the same sequence of frames as defined in Section 8.1. This stream becomes "half closed" to the client (Section 5.1) after the initial HEADERS frame is sent.

+

Once a client receives a PUSH_PROMISE frame and chooses to accept the pushed response, the client SHOULD NOT issue any requests for the promised response until after the promised stream has closed.

+

If the client determines, for any reason, that it does not wish to receive the pushed response from the server, or if the server takes too long to begin sending the promised response, the client can send an RST_STREAM frame, using either the CANCEL or REFUSED_STREAM codes, and referencing the pushed stream's identifier.

+

A client can use the SETTINGS_MAX_CONCURRENT_STREAMS setting to limit the number of responses that can be concurrently pushed by a server. Advertising a SETTINGS_MAX_CONCURRENT_STREAMS value of zero disables server push by preventing the server from creating the necessary streams. This does not prohibit a server from sending PUSH_PROMISE frames; clients need to reset any promised streams that are not wanted.

+

Clients receiving a pushed response MUST validate that either the server is authoritative (see Section 10.1), or the proxy that provided the pushed response is configured for the corresponding request. For example, a server that offers a certificate for only the example.com DNS-ID or Common Name is not permitted to push a response for https://www.example.org/doc.

+
+

The response for a PUSH_PROMISE stream begins with a HEADERS frame, which immediately puts the stream into the “half closed (remote)” state for the server and “half closed (local)” state for the client, and ends with a frame bearing END_STREAM, which places the stream in the "closed" state.

+
Note:
+
The client never sends a frame with the END_STREAM flag for a server push.
+
+
+
+
+
+

8.3 The CONNECT Method

+

In HTTP/1.x, the pseudo-method CONNECT ( + [RFC7231] , Section 4.3.6) is used to convert an HTTP connection into a tunnel to a remote host. CONNECT is primarily used with HTTP proxies to establish a TLS session with an origin server for the purposes of interacting with https resources.

+
+

In HTTP/2, the CONNECT method is used to establish a tunnel over a single HTTP/2 stream to a remote host, for similar purposes. The HTTP header field mapping works as defined in Request Header Fields (Section 8.1.2.3), with a few differences. Specifically:

+
    +
  • The :method header field is set to CONNECT.
  • +
  • The :scheme and :path header fields MUST be omitted.
  • +
  • The :authority header field contains the host and port to connect to (equivalent to the authority-form of the request-target of CONNECT requests, see + [RFC7230] , Section 5.3).
  • +
+
+

A CONNECT request that does not conform to these restrictions is malformed (Section 8.1.2.6).

+

A proxy that supports CONNECT establishes a TCP connection [TCP] to the server identified in the :authority header field. Once this connection is successfully established, the proxy sends a HEADERS frame containing a 2xx series status code to the client, as defined in + [RFC7231] , Section 4.3.6.

+

After the initial HEADERS frame sent by each peer, all subsequent DATA frames correspond to data sent on the TCP connection. The payload of any DATA frames sent by the client is transmitted by the proxy to the TCP server; data received from the TCP server is assembled into DATA frames by the proxy. Frame types other than DATA or stream management frames (RST_STREAM, WINDOW_UPDATE, and PRIORITY) MUST NOT be sent on a connected stream, and MUST be treated as a stream error (Section 5.4.2) if received.

+

The TCP connection can be closed by either peer. The END_STREAM flag on a DATA frame is treated as being equivalent to the TCP FIN bit. A client is expected to send a DATA frame with the END_STREAM flag set after receiving a frame bearing the END_STREAM flag. A proxy that receives a DATA frame with the END_STREAM flag set sends the attached data with the FIN bit set on the last TCP segment. A proxy that receives a TCP segment with the FIN bit set sends a DATA frame with the END_STREAM flag set. Note that the final TCP segment or DATA frame could be empty.

+

A TCP connection error is signaled with RST_STREAM. A proxy treats any error in the TCP connection, which includes receiving a TCP segment with the RST bit set, as a stream error (Section 5.4.2) of type CONNECT_ERROR. Correspondingly, a proxy MUST send a TCP segment with the RST bit set if it detects an error with the stream or the HTTP/2 connection.

+
+
+
+
+

10.1 Server Authority

+

HTTP/2 relies on the HTTP/1.1 definition of authority for determining whether a server is authoritative in providing a given response, see + [RFC7230] , Section 9.1. This relies on local name resolution for the "http" URI scheme, and the authenticated server identity for the "https" scheme (see + [RFC2818] , Section 3).

+
+
+

10.2 Cross-Protocol Attacks +

+

In a cross-protocol attack, an attacker causes a client to initiate a transaction in one protocol toward a server that understands a different protocol. An attacker might be able to cause the transaction to appear as valid transaction in the second protocol. In combination with the capabilities of the web context, this can be used to interact with poorly protected servers in private networks.

+

Completing a TLS handshake with an ALPN identifier for HTTP/2 can be considered sufficient protection against cross protocol attacks. ALPN provides a positive indication that a server is willing to proceed with HTTP/2, which prevents attacks on other TLS-based protocols.

+

The encryption in TLS makes it difficult for attackers to control the data which could be used in a cross-protocol attack on a cleartext protocol.

+

The cleartext version of HTTP/2 has minimal protection against cross-protocol attacks. The connection preface (Section 3.5) contains a string that is designed to confuse HTTP/1.1 servers, but no special protection is offered for other protocols. A server that is willing to ignore parts of an HTTP/1.1 request containing an Upgrade header field in addition to the client connection preface could be exposed to a cross-protocol attack.

+
+
+

10.3 Intermediary Encapsulation Attacks +

+

The HTTP/2 header field encoding allows the expression of names that are not valid field names in the Internet Message Syntax used by HTTP/1.1. Requests or responses containing invalid header field names MUST be treated as malformed (Section 8.1.2.6). An intermediary therefore cannot translate an HTTP/2 request or response containing an invalid field name into an HTTP/1.1 message.

+

Similarly, HTTP/2 allows header field values that are not valid. While most of the values that can be encoded will not alter header field parsing, carriage return (CR, ASCII 0xd), line feed (LF, ASCII 0xa), and the zero character (NUL, ASCII 0x0) might be exploited by an attacker if they are translated verbatim. Any request or response that contains a character not permitted in a header field value MUST be treated as malformed (Section 8.1.2.6). Valid characters are defined by the field-content ABNF rule in Section 3.2 of + [RFC7230] .

+
+
+

10.4 Cacheability of Pushed Responses +

+

Pushed responses do not have an explicit request from the client; the request is provided by the server in the PUSH_PROMISE frame.

+

Caching responses that are pushed is possible based on the guidance provided by the origin server in the Cache-Control header field. However, this can cause issues if a single server hosts more than one tenant. For example, a server might offer multiple users each a small portion of its URI space.

+

Where multiple tenants share space on the same server, that server MUST ensure that tenants are not able to push representations of resources that they do not have authority over. Failure to enforce this would allow a tenant to provide a representation that would be served out of cache, overriding the actual representation that the authoritative tenant provides.

+

Pushed responses for which an origin server is not authoritative (see Section 10.1) MUST NOT be used or cached.

+
+
+

10.5 Denial of Service Considerations

+

An HTTP/2 connection can demand a greater commitment of resources to operate than a HTTP/1.1 connection. The use of header compression and flow control depend on a commitment of resources for storing a greater amount of state. Settings for these features ensure that memory commitments for these features are strictly bounded.

+

The number of PUSH_PROMISE frames is not constrained in the same fashion. A client that accepts server push SHOULD limit the number of streams it allows to be in the "reserved (remote)" state. Excessive number of server push streams can be treated as a stream error (Section 5.4.2) of type ENHANCE_YOUR_CALM.

+

Processing capacity cannot be guarded as effectively as state capacity.

+

The SETTINGS frame can be abused to cause a peer to expend additional processing time. This might be done by pointlessly changing SETTINGS parameters, setting multiple undefined parameters, or changing the same setting multiple times in the same frame. WINDOW_UPDATE or PRIORITY frames can be abused to cause an unnecessary waste of resources.

+

Large numbers of small or empty frames can be abused to cause a peer to expend time processing frame headers. Note however that some uses are entirely legitimate, such as the sending of an empty DATA or CONTINUATION frame at the end of a stream.

+

Header compression also offers some opportunities to waste processing resources; see Section 7 of + [COMPRESSION] for more details on potential abuses.

+

Limits in SETTINGS parameters cannot be reduced instantaneously, which leaves an endpoint exposed to behavior from a peer that could exceed the new limits. In particular, immediately after establishing a connection, limits set by a server are not known to clients and could be exceeded without being an obvious protocol violation.

+

All these features - i.e., SETTINGS changes, small frames, header compression - have legitimate uses. These features become a burden only when they are used unnecessarily or to excess.

+

An endpoint that doesn't monitor this behavior exposes itself to a risk of denial of service attack. Implementations SHOULD track the use of these features and set limits on their use. An endpoint MAY treat activity that is suspicious as a connection error (Section 5.4.1) of type ENHANCE_YOUR_CALM.

+
+

10.5.2 CONNECT Issues

+

The CONNECT method can be used to create disproportionate load on an proxy, since stream creation is relatively inexpensive when compared to the creation and maintenance of a TCP connection. A proxy might also maintain some resources for a TCP connection beyond the closing of the stream that carries the CONNECT request, since the outgoing TCP connection remains in the TIME_WAIT state. A proxy therefore cannot rely on SETTINGS_MAX_CONCURRENT_STREAMS alone to limit the resources consumed by CONNECT requests.

+
+
+
+

10.6 Use of Compression +

+

Compression can allow an attacker to recover secret data when it is compressed in the same context as data under attacker control. HTTP/2 enables compression of header fields (Section 4.3); the following concerns also apply to the use of HTTP compressed content-codings ( + [RFC7231] , Section 3.1.2.1).

+

There are demonstrable attacks on compression that exploit the characteristics of the web (e.g., + [BREACH] ). The attacker induces multiple requests containing varying plaintext, observing the length of the resulting ciphertext in each, which reveals a shorter length when a guess about the secret is correct.

+

Implementations communicating on a secure channel MUST NOT compress content that includes both confidential and attacker-controlled data unless separate compression dictionaries are used for each source of data. Compression MUST NOT be used if the source of data cannot be reliably determined. Generic stream compression, such as that provided by TLS MUST NOT be used with HTTP/2 (see Section 9.2).

+

Further considerations regarding the compression of header fields are described in + [COMPRESSION] .

+
+
+

10.7 Use of Padding

+

Padding within HTTP/2 is not intended as a replacement for general purpose padding, such as might be provided by TLS [TLS12]. Redundant padding could even be counterproductive. Correct application can depend on having specific knowledge of the data that is being padded.

+

To mitigate attacks that rely on compression, disabling or limiting compression might be preferable to padding as a countermeasure.

+

Padding can be used to obscure the exact size of frame content, and is provided to mitigate specific attacks within HTTP. For example, attacks where compressed content includes both attacker-controlled plaintext and secret data (see for example, + [BREACH] ).

+

Use of padding can result in less protection than might seem immediately obvious. At best, padding only makes it more difficult for an attacker to infer length information by increasing the number of frames an attacker has to observe. Incorrectly implemented padding schemes can be easily defeated. In particular, randomized padding with a predictable distribution provides very little protection; similarly, padding payloads to a fixed size exposes information as payload sizes cross the fixed size boundary, which could be possible if an attacker can control plaintext.

+

Intermediaries SHOULD retain padding for DATA frames, but MAY drop padding for HEADERS and PUSH_PROMISE frames. A valid reason for an intermediary to change the amount of padding of frames is to improve the protections that padding provides.

+
+
+

10.8 Privacy Considerations +

+

Several characteristics of HTTP/2 provide an observer an opportunity to correlate actions of a single client or server over time. This includes the value of settings, the manner in which flow control windows are managed, the way priorities are allocated to streams, timing of reactions to stimulus, and handling of any features that are controlled by settings.

+

As far as this creates observable differences in behavior, they could be used as a basis for fingerprinting a specific client, as defined in Section 1.8 of + [HTML5] .

+

HTTP/2's preference for using a single TCP connection allows correlation of a user's activity on a site. If connections are reused for different origins, this allows tracking across those origins.

+

Because the PING and SETTINGS frames solicit immediate responses, they can be used by an endpoint to measure latency to their peer. This might have privacy implications in certain scenarios.

+
+
+
+

A string for identifying HTTP/2 is entered into the "Application Layer Protocol Negotiation (ALPN) Protocol IDs" registry established in + [TLS-ALPN] .

+

This document establishes a registry for frame types, settings, and error codes. These new registries are entered into a new "Hypertext Transfer Protocol (HTTP) 2 Parameters" section.

+

This document registers the HTTP2-Settings header field for use in HTTP; and the 421 (Misdirected Request) status code.

+

This document registers the PRI method for use in HTTP, to avoid collisions with the connection preface (Section 3.5).

+
+

11.1 Registration of HTTP/2 Identification Strings

+

This document creates two registrations for the identification of HTTP/2 in the "Application Layer Protocol Negotiation (ALPN) Protocol IDs" registry established in + [TLS-ALPN] .

+
+

The "h2" string identifies HTTP/2 when used over TLS:

+
Protocol:
+
HTTP/2 over TLS
Identification Sequence:
+
0x68 0x32 ("h2")
Specification:
+
This document
+
+
+
+

The "h2c" string identifies HTTP/2 when used over cleartext TCP:

+
Protocol:
+
HTTP/2 over TCP
Identification Sequence:
+
0x68 0x32 0x63 ("h2c")
Specification:
+
This document
+
+
+
+
+

11.2 Frame Type Registry

+

This document establishes a registry for HTTP/2 frame type codes. The "HTTP/2 Frame Type" registry manages an 8-bit space. The "HTTP/2 Frame Type" registry operates under either of the "IETF Review" or "IESG Approval" policies [RFC5226] for values between 0x00 and 0xef, with values between 0xf0 and 0xff being reserved for experimental use.

+
+

New entries in this registry require the following information:

+
Frame Type:
+
A name or label for the frame type.
Code:
+
The 8-bit code assigned to the frame type.
Specification:
+
A reference to a specification that includes a description of the frame layout, its semantics, and flags that the frame type uses, including any parts of the frame that are conditionally present based on the value of flags.
+
+
+

The entries in the following table are registered by this document.

+
+
+

11.3 Settings Registry

+

This document establishes a registry for HTTP/2 settings. The "HTTP/2 Settings" registry manages a 16-bit space. The "HTTP/2 Settings" registry operates under the "Expert Review" policy [RFC5226] for values in the range from 0x0000 to 0xefff, with values between and 0xf000 and 0xffff being reserved for experimental use.

+
+

New registrations are advised to provide the following information:

+
Name:
+
A symbolic name for the setting. Specifying a setting name is optional.
Code:
+
The 16-bit code assigned to the setting.
Initial Value:
+
An initial value for the setting.
Specification:
+
An optional reference to a specification that describes the use of the setting.
+
+
+

An initial set of setting registrations can be found in Section 6.5.2.

+
+
+

11.4 Error Code Registry

+

This document establishes a registry for HTTP/2 error codes. The "HTTP/2 Error Code" registry manages a 32-bit space. The "HTTP/2 Error Code" registry operates under the "Expert Review" policy [RFC5226].

+

Registrations for error codes are required to include a description of the error code. An expert reviewer is advised to examine new registrations for possible duplication with existing error codes. Use of existing registrations is to be encouraged, but not mandated.

+
+

New registrations are advised to provide the following information:

+
Name:
+
A name for the error code. Specifying an error code name is optional.
Code:
+
The 32-bit error code value.
Description:
+
A brief description of the error code semantics, longer if no detailed specification is provided.
Specification:
+
An optional reference for a specification that defines the error code.
+
+
+

The entries in the following table are registered by this document.

+
+
+

11.5 HTTP2-Settings Header Field Registration +

+
+

This section registers the HTTP2-Settings header field in the Permanent Message Header Field Registry [BCP90].

+
Header field name:
+
HTTP2-Settings
Applicable protocol:
+
http
Status:
+
standard
Author/Change controller:
+
IETF
Specification document(s):
+
Section 3.2.1 of this document
Related information:
+
This header field is only used by an HTTP/2 client for Upgrade-based negotiation.
+
+
+
+
+

11.6 PRI Method Registration +

+
+

This section registers the PRI method in the HTTP Method Registry ( + [RFC7231] , Section 8.1).

+
Method Name:
+
PRI
Safe
+
Yes
Idempotent
+
Yes
Specification document(s)
+
Section 3.5 of this document
Related information:
+
This method is never used by an actual client. This method will appear to be used when an HTTP/1.1 server or intermediary attempts to parse an HTTP/2 connection preface.
+
+
+
+
+
+
+

This document includes substantial input from the following individuals:

+
    +
  • Adam Langley, Wan-Teh Chang, Jim Morrison, Mark Nottingham, Alyssa Wilk, Costin Manolache, William Chan, Vitaliy Lvin, Joe Chan, Adam Barth, Ryan Hamilton, Gavin Peters, Kent Alstad, Kevin Lindsay, Paul Amer, Fan Yang, Jonathan Leighton (SPDY contributors).
  • +
  • Gabriel Montenegro and Willy Tarreau (Upgrade mechanism).
  • +
  • William Chan, Salvatore Loreto, Osama Mazahir, Gabriel Montenegro, Jitu Padhye, Roberto Peon, Rob Trace (Flow control).
  • +
  • Mike Bishop (Extensibility).
  • +
  • Mark Nottingham, Julian Reschke, James Snell, Jeff Pinner, Mike Bishop, Herve Ruellan (Substantial editorial contributions).
  • +
  • Kari Hurtta, Tatsuhiro Tsujikawa, Greg Wilkins, Poul-Henning Kamp, Jonathan Thackray.
  • +
  • Alexey Melnikov was an editor of this document during 2013.
  • +
  • A substantial proportion of Martin's contribution was supported by Microsoft during his employment there.
  • +
  • The Japanese HTTP/2 community provided an invaluable contribution, including a number of implementations, plus numerous technical and editorial contributions.
  • +
+
+
+

13.1 Normative References +

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
[COMPRESSION]Ruellan, H. and R. Peon, “HPACK - Header Compression for HTTP/2”, Internet-Draft draft-ietf-httpbis-header-compression-11 (work in progress), February 2015.
[COOKIE]Barth, A., “HTTP State Management Mechanism”, RFC 6265, April 2011, <http://www.rfc-editor.org/info/rfc6265>.
[FIPS186]NIST, “Digital Signature Standard (DSS)”, FIPS PUB 186-4, July 2013, <http://dx.doi.org/10.6028/NIST.FIPS.186-4>.
[RFC2119]Bradner, S., “Key words for use in RFCs to Indicate Requirement Levels”, BCP 14, RFC 2119, March 1997, <http://www.rfc-editor.org/info/rfc2119>.
[RFC2818]Rescorla, E., “HTTP Over TLS”, RFC 2818, May 2000, <http://www.rfc-editor.org/info/rfc2818>.
[RFC3986]Berners-Lee, T., Fielding, R., and L. Masinter, “Uniform Resource Identifier (URI): Generic Syntax”, STD 66, RFC 3986, January 2005, <http://www.rfc-editor.org/info/rfc3986>.
[RFC4648]Josefsson, S., “The Base16, Base32, and Base64 Data Encodings”, RFC 4648, October 2006, <http://www.rfc-editor.org/info/rfc4648>.
[RFC5226]Narten, T. and H. Alvestrand, “Guidelines for Writing an IANA Considerations Section in RFCs”, BCP 26, RFC 5226, May 2008, <http://www.rfc-editor.org/info/rfc5226>.
[RFC5234]Crocker, D. and P. Overell, “Augmented BNF for Syntax Specifications: ABNF”, STD 68, RFC 5234, January 2008, <http://www.rfc-editor.org/info/rfc5234>.
[RFC7230]Fielding, R., Ed. and J. Reschke, Ed., “Hypertext Transfer Protocol (HTTP/1.1): Message Syntax and Routing”, RFC 7230, June 2014, <http://www.rfc-editor.org/info/rfc7230>.
[RFC7231]Fielding, R., Ed. and J. Reschke, Ed., “Hypertext Transfer Protocol (HTTP/1.1): Semantics and Content”, RFC 7231, June 2014, <http://www.rfc-editor.org/info/rfc7231>.
[RFC7232]Fielding, R., Ed. and J. Reschke, Ed., “Hypertext Transfer Protocol (HTTP/1.1): Conditional Requests”, RFC 7232, June 2014, <http://www.rfc-editor.org/info/rfc7232>.
[RFC7233]Fielding, R., Ed., Lafon, Y., Ed., and J. Reschke, Ed., “Hypertext Transfer Protocol (HTTP/1.1): Range Requests”, RFC 7233, June 2014, <http://www.rfc-editor.org/info/rfc7233>.
[RFC7234]Fielding, R., Ed., Nottingham, M., Ed., and J. Reschke, Ed., “Hypertext Transfer Protocol (HTTP/1.1): Caching”, RFC 7234, June 2014, <http://www.rfc-editor.org/info/rfc7234>.
[RFC7235]Fielding, R., Ed. and J. Reschke, Ed., “Hypertext Transfer Protocol (HTTP/1.1): Authentication”, RFC 7235, June 2014, <http://www.rfc-editor.org/info/rfc7235>.
[TCP]Postel, J., “Transmission Control Protocol”, STD 7, RFC 793, September 1981, <http://www.rfc-editor.org/info/rfc793>.
[TLS-ALPN]Friedl, S., Popov, A., Langley, A., and E. Stephan, “Transport Layer Security (TLS) Application-Layer Protocol Negotiation Extension”, RFC 7301, July 2014, <http://www.rfc-editor.org/info/rfc7301>.
[TLS-ECDHE]Rescorla, E., “TLS Elliptic Curve Cipher Suites with SHA-256/384 and AES Galois Counter Mode (GCM)”, RFC 5289, August 2008, <http://www.rfc-editor.org/info/rfc5289>.
[TLS-EXT]Eastlake, D., “Transport Layer Security (TLS) Extensions: Extension Definitions”, RFC 6066, January 2011, <http://www.rfc-editor.org/info/rfc6066>.
[TLS12]Dierks, T. and E. Rescorla, “The Transport Layer Security (TLS) Protocol Version 1.2”, RFC 5246, August 2008, <http://www.rfc-editor.org/info/rfc5246>.
+

13.2 Informative References +

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
[ALT-SVC]Nottingham, M., McManus, P., and J. Reschke, “HTTP Alternative Services”, Internet-Draft draft-ietf-httpbis-alt-svc-06 (work in progress), February 2015.
[BCP90]Klyne, G., Nottingham, M., and J. Mogul, “Registration Procedures for Message Header Fields”, BCP 90, RFC 3864, September 2004, <http://www.rfc-editor.org/info/rfc3864>.
[BREACH]Gluck, Y., Harris, N., and A. Prado, “BREACH: Reviving the CRIME Attack”, July 2013, <http://breachattack.com/resources/BREACH%20-%20SSL,%20gone%20in%2030%20seconds.pdf>.
[HTML5]Hickson, I., Berjon, R., Faulkner, S., Leithead, T., Doyle Navara, E., O'Connor, E., and S. Pfeiffer, “HTML5”, W3C Recommendation REC-html5-20141028, October 2014, <http://www.w3.org/TR/2014/REC-html5-20141028/>. +
Latest version available at <http://www.w3.org/TR/html5/>.
[RFC3749]Hollenbeck, S., “Transport Layer Security Protocol Compression Methods”, RFC 3749, May 2004, <http://www.rfc-editor.org/info/rfc3749>.
[RFC4492]Blake-Wilson, S., Bolyard, N., Gupta, V., Hawk, C., and B. Moeller, “Elliptic Curve Cryptography (ECC) Cipher Suites for Transport Layer Security (TLS)”, RFC 4492, May 2006, <http://www.rfc-editor.org/info/rfc4492>.
[RFC6585]Nottingham, M. and R. Fielding, “Additional HTTP Status Codes”, RFC 6585, April 2012, <http://www.rfc-editor.org/info/rfc6585>.
[RFC7323]Borman, D., Braden, B., Jacobson, V., and R. Scheffenegger, “TCP Extensions for High Performance”, RFC 7323, September 2014, <http://www.rfc-editor.org/info/rfc7323>.
[TALKING]Huang, L-S., Chen, E., Barth, A., Rescorla, E., and C. Jackson, “Talking to Yourself for Fun and Profit”, 2011, <http://w2spconf.com/2011/papers/websocket.pdf>.
[TLSBCP]Sheffer, Y., Holz, R., and P. Saint-Andre, “Recommendations for Secure Use of TLS and DTLS”, Internet-Draft draft-ietf-uta-tls-bcp-08 (work in progress), December 2014.
+
+
+

An HTTP/2 implementation MAY treat the negotiation of any of the following cipher suites with TLS 1.2 as a connection error (Section 5.4.1) of type INADEQUATE_SECURITY: TLS_NULL_WITH_NULL_NULL, TLS_RSA_WITH_NULL_MD5, TLS_RSA_WITH_NULL_SHA, TLS_RSA_EXPORT_WITH_RC4_40_MD5, TLS_RSA_WITH_RC4_128_MD5, TLS_RSA_WITH_RC4_128_SHA, TLS_RSA_EXPORT_WITH_RC2_CBC_40_MD5, TLS_RSA_WITH_IDEA_CBC_SHA, TLS_RSA_EXPORT_WITH_DES40_CBC_SHA, TLS_RSA_WITH_DES_CBC_SHA, TLS_RSA_WITH_3DES_EDE_CBC_SHA, TLS_DH_DSS_EXPORT_WITH_DES40_CBC_SHA, TLS_DH_DSS_WITH_DES_CBC_SHA, TLS_DH_DSS_WITH_3DES_EDE_CBC_SHA, TLS_DH_RSA_EXPORT_WITH_DES40_CBC_SHA, TLS_DH_RSA_WITH_DES_CBC_SHA, TLS_DH_RSA_WITH_3DES_EDE_CBC_SHA, TLS_DHE_DSS_EXPORT_WITH_DES40_CBC_SHA, TLS_DHE_DSS_WITH_DES_CBC_SHA, TLS_DHE_DSS_WITH_3DES_EDE_CBC_SHA, TLS_DHE_RSA_EXPORT_WITH_DES40_CBC_SHA, TLS_DHE_RSA_WITH_DES_CBC_SHA, TLS_DHE_RSA_WITH_3DES_EDE_CBC_SHA, TLS_DH_anon_EXPORT_WITH_RC4_40_MD5, TLS_DH_anon_WITH_RC4_128_MD5, TLS_DH_anon_EXPORT_WITH_DES40_CBC_SHA, TLS_DH_anon_WITH_DES_CBC_SHA, TLS_DH_anon_WITH_3DES_EDE_CBC_SHA, TLS_KRB5_WITH_DES_CBC_SHA, TLS_KRB5_WITH_3DES_EDE_CBC_SHA, TLS_KRB5_WITH_RC4_128_SHA, TLS_KRB5_WITH_IDEA_CBC_SHA, TLS_KRB5_WITH_DES_CBC_MD5, TLS_KRB5_WITH_3DES_EDE_CBC_MD5, TLS_KRB5_WITH_RC4_128_MD5, TLS_KRB5_WITH_IDEA_CBC_MD5, TLS_KRB5_EXPORT_WITH_DES_CBC_40_SHA, TLS_KRB5_EXPORT_WITH_RC2_CBC_40_SHA, TLS_KRB5_EXPORT_WITH_RC4_40_SHA, TLS_KRB5_EXPORT_WITH_DES_CBC_40_MD5, TLS_KRB5_EXPORT_WITH_RC2_CBC_40_MD5, TLS_KRB5_EXPORT_WITH_RC4_40_MD5, TLS_PSK_WITH_NULL_SHA, TLS_DHE_PSK_WITH_NULL_SHA, TLS_RSA_PSK_WITH_NULL_SHA, TLS_RSA_WITH_AES_128_CBC_SHA, TLS_DH_DSS_WITH_AES_128_CBC_SHA, TLS_DH_RSA_WITH_AES_128_CBC_SHA, TLS_DHE_DSS_WITH_AES_128_CBC_SHA, TLS_DHE_RSA_WITH_AES_128_CBC_SHA, TLS_DH_anon_WITH_AES_128_CBC_SHA, TLS_RSA_WITH_AES_256_CBC_SHA, TLS_DH_DSS_WITH_AES_256_CBC_SHA, TLS_DH_RSA_WITH_AES_256_CBC_SHA, TLS_DHE_DSS_WITH_AES_256_CBC_SHA, TLS_DHE_RSA_WITH_AES_256_CBC_SHA, TLS_DH_anon_WITH_AES_256_CBC_SHA, TLS_RSA_WITH_NULL_SHA256, TLS_RSA_WITH_AES_128_CBC_SHA256, TLS_RSA_WITH_AES_256_CBC_SHA256, TLS_DH_DSS_WITH_AES_128_CBC_SHA256, TLS_DH_RSA_WITH_AES_128_CBC_SHA256, TLS_DHE_DSS_WITH_AES_128_CBC_SHA256, TLS_RSA_WITH_CAMELLIA_128_CBC_SHA, TLS_DH_DSS_WITH_CAMELLIA_128_CBC_SHA, TLS_DH_RSA_WITH_CAMELLIA_128_CBC_SHA, TLS_DHE_DSS_WITH_CAMELLIA_128_CBC_SHA, TLS_DHE_RSA_WITH_CAMELLIA_128_CBC_SHA, TLS_DH_anon_WITH_CAMELLIA_128_CBC_SHA, TLS_DHE_RSA_WITH_AES_128_CBC_SHA256, TLS_DH_DSS_WITH_AES_256_CBC_SHA256, TLS_DH_RSA_WITH_AES_256_CBC_SHA256, TLS_DHE_DSS_WITH_AES_256_CBC_SHA256, TLS_DHE_RSA_WITH_AES_256_CBC_SHA256, TLS_DH_anon_WITH_AES_128_CBC_SHA256, TLS_DH_anon_WITH_AES_256_CBC_SHA256, TLS_RSA_WITH_CAMELLIA_256_CBC_SHA, TLS_DH_DSS_WITH_CAMELLIA_256_CBC_SHA, TLS_DH_RSA_WITH_CAMELLIA_256_CBC_SHA, TLS_DHE_DSS_WITH_CAMELLIA_256_CBC_SHA, TLS_DHE_RSA_WITH_CAMELLIA_256_CBC_SHA, TLS_DH_anon_WITH_CAMELLIA_256_CBC_SHA, TLS_PSK_WITH_RC4_128_SHA, TLS_PSK_WITH_3DES_EDE_CBC_SHA, TLS_PSK_WITH_AES_128_CBC_SHA, TLS_PSK_WITH_AES_256_CBC_SHA, TLS_DHE_PSK_WITH_RC4_128_SHA, TLS_DHE_PSK_WITH_3DES_EDE_CBC_SHA, TLS_DHE_PSK_WITH_AES_128_CBC_SHA, TLS_DHE_PSK_WITH_AES_256_CBC_SHA, TLS_RSA_PSK_WITH_RC4_128_SHA, TLS_RSA_PSK_WITH_3DES_EDE_CBC_SHA, TLS_RSA_PSK_WITH_AES_128_CBC_SHA, TLS_RSA_PSK_WITH_AES_256_CBC_SHA, TLS_RSA_WITH_SEED_CBC_SHA, TLS_DH_DSS_WITH_SEED_CBC_SHA, TLS_DH_RSA_WITH_SEED_CBC_SHA, TLS_DHE_DSS_WITH_SEED_CBC_SHA, TLS_DHE_RSA_WITH_SEED_CBC_SHA, TLS_DH_anon_WITH_SEED_CBC_SHA, TLS_RSA_WITH_AES_128_GCM_SHA256, TLS_RSA_WITH_AES_256_GCM_SHA384, TLS_DH_RSA_WITH_AES_128_GCM_SHA256, TLS_DH_RSA_WITH_AES_256_GCM_SHA384, TLS_DH_DSS_WITH_AES_128_GCM_SHA256, TLS_DH_DSS_WITH_AES_256_GCM_SHA384, TLS_DH_anon_WITH_AES_128_GCM_SHA256, TLS_DH_anon_WITH_AES_256_GCM_SHA384, TLS_PSK_WITH_AES_128_GCM_SHA256, TLS_PSK_WITH_AES_256_GCM_SHA384, TLS_RSA_PSK_WITH_AES_128_GCM_SHA256, TLS_RSA_PSK_WITH_AES_256_GCM_SHA384, TLS_PSK_WITH_AES_128_CBC_SHA256, TLS_PSK_WITH_AES_256_CBC_SHA384, TLS_PSK_WITH_NULL_SHA256, TLS_PSK_WITH_NULL_SHA384, TLS_DHE_PSK_WITH_AES_128_CBC_SHA256, TLS_DHE_PSK_WITH_AES_256_CBC_SHA384, TLS_DHE_PSK_WITH_NULL_SHA256, TLS_DHE_PSK_WITH_NULL_SHA384, TLS_RSA_PSK_WITH_AES_128_CBC_SHA256, TLS_RSA_PSK_WITH_AES_256_CBC_SHA384, TLS_RSA_PSK_WITH_NULL_SHA256, TLS_RSA_PSK_WITH_NULL_SHA384, TLS_RSA_WITH_CAMELLIA_128_CBC_SHA256, TLS_DH_DSS_WITH_CAMELLIA_128_CBC_SHA256, TLS_DH_RSA_WITH_CAMELLIA_128_CBC_SHA256, TLS_DHE_DSS_WITH_CAMELLIA_128_CBC_SHA256, TLS_DHE_RSA_WITH_CAMELLIA_128_CBC_SHA256, TLS_DH_anon_WITH_CAMELLIA_128_CBC_SHA256, TLS_RSA_WITH_CAMELLIA_256_CBC_SHA256, TLS_DH_DSS_WITH_CAMELLIA_256_CBC_SHA256, TLS_DH_RSA_WITH_CAMELLIA_256_CBC_SHA256, TLS_DHE_DSS_WITH_CAMELLIA_256_CBC_SHA256, TLS_DHE_RSA_WITH_CAMELLIA_256_CBC_SHA256, TLS_DH_anon_WITH_CAMELLIA_256_CBC_SHA256, TLS_EMPTY_RENEGOTIATION_INFO_SCSV, TLS_ECDH_ECDSA_WITH_NULL_SHA, TLS_ECDH_ECDSA_WITH_RC4_128_SHA, TLS_ECDH_ECDSA_WITH_3DES_EDE_CBC_SHA, TLS_ECDH_ECDSA_WITH_AES_128_CBC_SHA, TLS_ECDH_ECDSA_WITH_AES_256_CBC_SHA, TLS_ECDHE_ECDSA_WITH_NULL_SHA, TLS_ECDHE_ECDSA_WITH_RC4_128_SHA, TLS_ECDHE_ECDSA_WITH_3DES_EDE_CBC_SHA, TLS_ECDHE_ECDSA_WITH_AES_128_CBC_SHA, TLS_ECDHE_ECDSA_WITH_AES_256_CBC_SHA, TLS_ECDH_RSA_WITH_NULL_SHA, TLS_ECDH_RSA_WITH_RC4_128_SHA, TLS_ECDH_RSA_WITH_3DES_EDE_CBC_SHA, TLS_ECDH_RSA_WITH_AES_128_CBC_SHA, TLS_ECDH_RSA_WITH_AES_256_CBC_SHA, TLS_ECDHE_RSA_WITH_NULL_SHA, TLS_ECDHE_RSA_WITH_RC4_128_SHA, TLS_ECDHE_RSA_WITH_3DES_EDE_CBC_SHA, TLS_ECDHE_RSA_WITH_AES_128_CBC_SHA, TLS_ECDHE_RSA_WITH_AES_256_CBC_SHA, TLS_ECDH_anon_WITH_NULL_SHA, TLS_ECDH_anon_WITH_RC4_128_SHA, TLS_ECDH_anon_WITH_3DES_EDE_CBC_SHA, TLS_ECDH_anon_WITH_AES_128_CBC_SHA, TLS_ECDH_anon_WITH_AES_256_CBC_SHA, TLS_SRP_SHA_WITH_3DES_EDE_CBC_SHA, TLS_SRP_SHA_RSA_WITH_3DES_EDE_CBC_SHA, TLS_SRP_SHA_DSS_WITH_3DES_EDE_CBC_SHA, TLS_SRP_SHA_WITH_AES_128_CBC_SHA, TLS_SRP_SHA_RSA_WITH_AES_128_CBC_SHA, TLS_SRP_SHA_DSS_WITH_AES_128_CBC_SHA, TLS_SRP_SHA_WITH_AES_256_CBC_SHA, TLS_SRP_SHA_RSA_WITH_AES_256_CBC_SHA, TLS_SRP_SHA_DSS_WITH_AES_256_CBC_SHA, TLS_ECDHE_ECDSA_WITH_AES_128_CBC_SHA256, TLS_ECDHE_ECDSA_WITH_AES_256_CBC_SHA384, TLS_ECDH_ECDSA_WITH_AES_128_CBC_SHA256, TLS_ECDH_ECDSA_WITH_AES_256_CBC_SHA384, TLS_ECDHE_RSA_WITH_AES_128_CBC_SHA256, TLS_ECDHE_RSA_WITH_AES_256_CBC_SHA384, TLS_ECDH_RSA_WITH_AES_128_CBC_SHA256, TLS_ECDH_RSA_WITH_AES_256_CBC_SHA384, TLS_ECDH_ECDSA_WITH_AES_128_GCM_SHA256, TLS_ECDH_ECDSA_WITH_AES_256_GCM_SHA384, TLS_ECDH_RSA_WITH_AES_128_GCM_SHA256, TLS_ECDH_RSA_WITH_AES_256_GCM_SHA384, TLS_ECDHE_PSK_WITH_RC4_128_SHA, TLS_ECDHE_PSK_WITH_3DES_EDE_CBC_SHA, TLS_ECDHE_PSK_WITH_AES_128_CBC_SHA, TLS_ECDHE_PSK_WITH_AES_256_CBC_SHA, TLS_ECDHE_PSK_WITH_AES_128_CBC_SHA256, TLS_ECDHE_PSK_WITH_AES_256_CBC_SHA384, TLS_ECDHE_PSK_WITH_NULL_SHA, TLS_ECDHE_PSK_WITH_NULL_SHA256, TLS_ECDHE_PSK_WITH_NULL_SHA384, TLS_RSA_WITH_ARIA_128_CBC_SHA256, TLS_RSA_WITH_ARIA_256_CBC_SHA384, TLS_DH_DSS_WITH_ARIA_128_CBC_SHA256, TLS_DH_DSS_WITH_ARIA_256_CBC_SHA384, TLS_DH_RSA_WITH_ARIA_128_CBC_SHA256, TLS_DH_RSA_WITH_ARIA_256_CBC_SHA384, TLS_DHE_DSS_WITH_ARIA_128_CBC_SHA256, TLS_DHE_DSS_WITH_ARIA_256_CBC_SHA384, TLS_DHE_RSA_WITH_ARIA_128_CBC_SHA256, TLS_DHE_RSA_WITH_ARIA_256_CBC_SHA384, TLS_DH_anon_WITH_ARIA_128_CBC_SHA256, TLS_DH_anon_WITH_ARIA_256_CBC_SHA384, TLS_ECDHE_ECDSA_WITH_ARIA_128_CBC_SHA256, TLS_ECDHE_ECDSA_WITH_ARIA_256_CBC_SHA384, TLS_ECDH_ECDSA_WITH_ARIA_128_CBC_SHA256, TLS_ECDH_ECDSA_WITH_ARIA_256_CBC_SHA384, TLS_ECDHE_RSA_WITH_ARIA_128_CBC_SHA256, TLS_ECDHE_RSA_WITH_ARIA_256_CBC_SHA384, TLS_ECDH_RSA_WITH_ARIA_128_CBC_SHA256, TLS_ECDH_RSA_WITH_ARIA_256_CBC_SHA384, TLS_RSA_WITH_ARIA_128_GCM_SHA256, TLS_RSA_WITH_ARIA_256_GCM_SHA384, TLS_DH_RSA_WITH_ARIA_128_GCM_SHA256, TLS_DH_RSA_WITH_ARIA_256_GCM_SHA384, TLS_DH_DSS_WITH_ARIA_128_GCM_SHA256, TLS_DH_DSS_WITH_ARIA_256_GCM_SHA384, TLS_DH_anon_WITH_ARIA_128_GCM_SHA256, TLS_DH_anon_WITH_ARIA_256_GCM_SHA384, TLS_ECDH_ECDSA_WITH_ARIA_128_GCM_SHA256, TLS_ECDH_ECDSA_WITH_ARIA_256_GCM_SHA384, TLS_ECDH_RSA_WITH_ARIA_128_GCM_SHA256, TLS_ECDH_RSA_WITH_ARIA_256_GCM_SHA384, TLS_PSK_WITH_ARIA_128_CBC_SHA256, TLS_PSK_WITH_ARIA_256_CBC_SHA384, TLS_DHE_PSK_WITH_ARIA_128_CBC_SHA256, TLS_DHE_PSK_WITH_ARIA_256_CBC_SHA384, TLS_RSA_PSK_WITH_ARIA_128_CBC_SHA256, TLS_RSA_PSK_WITH_ARIA_256_CBC_SHA384, TLS_PSK_WITH_ARIA_128_GCM_SHA256, TLS_PSK_WITH_ARIA_256_GCM_SHA384, TLS_RSA_PSK_WITH_ARIA_128_GCM_SHA256, TLS_RSA_PSK_WITH_ARIA_256_GCM_SHA384, TLS_ECDHE_PSK_WITH_ARIA_128_CBC_SHA256, TLS_ECDHE_PSK_WITH_ARIA_256_CBC_SHA384, TLS_ECDHE_ECDSA_WITH_CAMELLIA_128_CBC_SHA256, TLS_ECDHE_ECDSA_WITH_CAMELLIA_256_CBC_SHA384, TLS_ECDH_ECDSA_WITH_CAMELLIA_128_CBC_SHA256, TLS_ECDH_ECDSA_WITH_CAMELLIA_256_CBC_SHA384, TLS_ECDHE_RSA_WITH_CAMELLIA_128_CBC_SHA256, TLS_ECDHE_RSA_WITH_CAMELLIA_256_CBC_SHA384, TLS_ECDH_RSA_WITH_CAMELLIA_128_CBC_SHA256, TLS_ECDH_RSA_WITH_CAMELLIA_256_CBC_SHA384, TLS_RSA_WITH_CAMELLIA_128_GCM_SHA256, TLS_RSA_WITH_CAMELLIA_256_GCM_SHA384, TLS_DH_RSA_WITH_CAMELLIA_128_GCM_SHA256, TLS_DH_RSA_WITH_CAMELLIA_256_GCM_SHA384, TLS_DH_DSS_WITH_CAMELLIA_128_GCM_SHA256, TLS_DH_DSS_WITH_CAMELLIA_256_GCM_SHA384, TLS_DH_anon_WITH_CAMELLIA_128_GCM_SHA256, TLS_DH_anon_WITH_CAMELLIA_256_GCM_SHA384, TLS_ECDH_ECDSA_WITH_CAMELLIA_128_GCM_SHA256, TLS_ECDH_ECDSA_WITH_CAMELLIA_256_GCM_SHA384, TLS_ECDH_RSA_WITH_CAMELLIA_128_GCM_SHA256, TLS_ECDH_RSA_WITH_CAMELLIA_256_GCM_SHA384, TLS_PSK_WITH_CAMELLIA_128_GCM_SHA256, TLS_PSK_WITH_CAMELLIA_256_GCM_SHA384, TLS_RSA_PSK_WITH_CAMELLIA_128_GCM_SHA256, TLS_RSA_PSK_WITH_CAMELLIA_256_GCM_SHA384, TLS_PSK_WITH_CAMELLIA_128_CBC_SHA256, TLS_PSK_WITH_CAMELLIA_256_CBC_SHA384, TLS_DHE_PSK_WITH_CAMELLIA_128_CBC_SHA256, TLS_DHE_PSK_WITH_CAMELLIA_256_CBC_SHA384, TLS_RSA_PSK_WITH_CAMELLIA_128_CBC_SHA256, TLS_RSA_PSK_WITH_CAMELLIA_256_CBC_SHA384, TLS_ECDHE_PSK_WITH_CAMELLIA_128_CBC_SHA256, TLS_ECDHE_PSK_WITH_CAMELLIA_256_CBC_SHA384, TLS_RSA_WITH_AES_128_CCM, TLS_RSA_WITH_AES_256_CCM, TLS_RSA_WITH_AES_128_CCM_8, TLS_RSA_WITH_AES_256_CCM_8, TLS_PSK_WITH_AES_128_CCM, TLS_PSK_WITH_AES_256_CCM, TLS_PSK_WITH_AES_128_CCM_8, TLS_PSK_WITH_AES_256_CCM_8.

+
Note:
+
This list was assembled from the set of registered TLS cipher suites at the time of writing. This list includes those cipher suites that do not offer an ephemeral key exchange and those that are based on the TLS null, stream or block cipher type (as defined in Section 6.2.3 of + [TLS12] ). Additional cipher suites with these properties could be defined; these would not be explicitly prohibited.
+
+
+
+
+

This section is to be removed by RFC Editor before publication.

+
+

B.3 Since draft-ietf-httpbis-http2-13

+

Pseudo-header fields are now required to appear strictly before regular ones.

+

Restored 1xx series status codes, except 101.

+

Changed frame length field 24-bits. Expanded frame header to 9 octets. Added a setting to limit the damage.

+

Added a setting to advise peers of header set size limits.

+

Removed segments.

+

Made non-semantic-bearing HEADERS frames illegal in the HTTP mapping.

+
+
+

B.4 Since draft-ietf-httpbis-http2-12

+

Restored extensibility options.

+

Restricting TLS cipher suites to AEAD only.

+

Removing Content-Encoding requirements.

+

Permitting the use of PRIORITY after stream close.

+

Removed ALTSVC frame.

+

Removed BLOCKED frame.

+

Reducing the maximum padding size to 256 octets; removing padding from CONTINUATION frames.

+

Removed per-frame GZIP compression.

+
+
+

B.6 Since draft-ietf-httpbis-http2-10

+

Changed "connection header" to "connection preface" to avoid confusion.

+

Added dependency-based stream prioritization.

+

Added "h2c" identifier to distinguish between cleartext and secured HTTP/2.

+

Adding missing padding to PUSH_PROMISE.

+

Integrate ALTSVC frame and supporting text.

+

Dropping requirement on "deflate" Content-Encoding.

+

Improving security considerations around use of compression.

+
+
+

B.7 Since draft-ietf-httpbis-http2-09

+

Adding padding for data frames.

+

Renumbering frame types, error codes, and settings.

+

Adding INADEQUATE_SECURITY error code.

+

Updating TLS usage requirements to 1.2; forbidding TLS compression.

+

Removing extensibility for frames and settings.

+

Changing setting identifier size.

+

Removing the ability to disable flow control.

+

Changing the protocol identification token to "h2".

+

Changing the use of :authority to make it optional and to allow userinfo in non-HTTP cases.

+

Allowing split on 0x0 for Cookie.

+

Reserved PRI method in HTTP/1.1 to avoid possible future collisions.

+
+
+

B.10 Since draft-ietf-httpbis-http2-06

+

Adding definition for CONNECT method.

+

Constraining the use of push to safe, cacheable methods with no request body.

+

Changing from :host to :authority to remove any potential confusion.

+

Adding setting for header compression table size.

+

Adding settings acknowledgement.

+

Removing unnecessary and potentially problematic flags from CONTINUATION.

+

Added denial of service considerations.

+
+
+

B.12 Since draft-ietf-httpbis-http2-04

+

Added CONTINUATION frame for HEADERS and PUSH_PROMISE.

+

PUSH_PROMISE is no longer implicitly prohibited if SETTINGS_MAX_CONCURRENT_STREAMS is zero.

+

Push expanded to allow all safe methods without a request body.

+

Clarified the use of HTTP header fields in requests and responses. Prohibited HTTP/1.1 hop-by-hop header fields.

+

Requiring that intermediaries not forward requests with missing or illegal routing :-headers.

+

Clarified requirements around handling different frames after stream close, stream reset and GOAWAY.

+

Added more specific prohibitions for sending of different frame types in various stream states.

+

Making the last received setting value the effective value.

+

Clarified requirements on TLS version, extension and ciphers.

+
+
+

B.13 Since draft-ietf-httpbis-http2-03

+

Committed major restructuring atrocities.

+

Added reference to first header compression draft.

+

Added more formal description of frame lifecycle.

+

Moved END_STREAM (renamed from FINAL) back to HEADERS/DATA.

+

Removed HEADERS+PRIORITY, added optional priority to HEADERS frame.

+

Added PRIORITY frame.

+
+
+

B.14 Since draft-ietf-httpbis-http2-02

+

Added continuations to frames carrying header blocks.

+

Replaced use of "session" with "connection" to avoid confusion with other HTTP stateful concepts, like cookies.

+

Removed "message".

+

Switched to TLS ALPN from NPN.

+

Editorial changes.

+
+
+

B.15 Since draft-ietf-httpbis-http2-01

+

Added IANA considerations section for frame types, error codes and settings.

+

Removed data frame compression.

+

Added PUSH_PROMISE.

+

Added globally applicable flags to framing.

+

Removed zlib-based header compression mechanism.

+

Updated references.

+

Clarified stream identifier reuse.

+

Removed CREDENTIALS frame and associated mechanisms.

+

Added advice against naive implementation of flow control.

+

Added session header section.

+

Restructured frame header. Removed distinction between data and control frames.

+

Altered flow control properties to include session-level limits.

+

Added note on cacheability of pushed resources and multiple tenant servers.

+

Changed protocol label form based on discussions.

+
+
+
+

Mike Belshe +
Twist +
Email: mbelshe@chromium.org

+

Roberto Peon +
Google, Inc +
Email: fenix@google.com

+

Martin Thomson (editor) +
Mozilla +
331 E Evelyn Street +
Mountain View, CA 94041 +
US +
Email: martin.thomson@gmail.com

+
+ +
+
\ No newline at end of file diff --git a/test/test-pages/http2-spec/source.html b/test/test-pages/http2-spec/source.html new file mode 100644 index 0000000..6a44fab --- /dev/null +++ b/test/test-pages/http2-spec/source.html @@ -0,0 +1,4075 @@ + + + + + + Hypertext Transfer Protocol version 2 + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
HTTPbis Working GroupM. Belshe
Internet-DraftTwist
Intended status: Standards TrackR. Peon
Expires: November 5, 2015Google, Inc
M. Thomson, Editor
Mozilla
May 4, 2015
+

Hypertext Transfer Protocol version 2 +
draft-ietf-httpbis-http2-latest

+

Abstract

+

This specification describes an optimized expression of the semantics of the Hypertext Transfer Protocol (HTTP). HTTP/2 enables a more efficient use of network resources and a reduced perception of latency by introducing header field compression and allowing multiple concurrent exchanges on the same connection. It also introduces unsolicited push of representations from servers to clients.

+

This specification is an alternative to, but does not obsolete, the HTTP/1.1 message syntax. HTTP's existing semantics remain unchanged.

+

Editorial Note (To be removed by RFC Editor)

+

Discussion of this draft takes place on the HTTPBIS working group mailing list (ietf-http-wg@w3.org), which is archived at <https://lists.w3.org/Archives/Public/ietf-http-wg/>.

+

Working Group information can be found at <https://tools.ietf.org/wg/httpbis/>; that specific to HTTP/2 are at <https://http2.github.io/>.

+

The changes in this draft are summarized in Appendix B.

+
+

Status of This Memo

+

This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79.

+

Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet-Drafts is at http://datatracker.ietf.org/drafts/current/.

+

Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as “work in progress”.

+

This Internet-Draft will expire on November 5, 2015.

+
+
+

Copyright Notice

+

Copyright © 2015 IETF Trust and the persons identified as the document authors. All rights reserved.

+

This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (http://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Code Components extracted from this document must include Simplified BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Simplified BSD License.

+
+
+
+

Table of Contents

+ +
+ +
+

1. Introduction

+
+

The Hypertext Transfer Protocol (HTTP) is a wildly successful protocol. However, how HTTP/1.1 uses the underlying transport ( + + [RFC7230] + , Section 6) has several characteristics that have a negative overall effect on application performance today.

+
+
+

In particular, HTTP/1.0 allowed only one request to be outstanding at a time on a given TCP connection. HTTP/1.1 added request pipelining, but this only partially addressed request concurrency and still suffers from head-of-line blocking. Therefore, HTTP/1.0 and HTTP/1.1 clients that need to make many requests use multiple connections to a server in order to achieve concurrency and thereby reduce latency.

+
+
+

Furthermore, HTTP header fields are often repetitive and verbose, causing unnecessary network traffic, as well as causing the initial TCP + [TCP] congestion window to quickly fill. This can result in excessive latency when multiple requests are made on a new TCP connection.

+
+
+

HTTP/2 addresses these issues by defining an optimized mapping of HTTP's semantics to an underlying connection. Specifically, it allows interleaving of request and response messages on the same connection and uses an efficient coding for HTTP header fields. It also allows prioritization of requests, letting more important requests complete more quickly, further improving performance.

+
+
+

The resulting protocol is more friendly to the network, because fewer TCP connections can be used in comparison to HTTP/1.x. This means less competition with other flows, and longer-lived connections, which in turn leads to better utilization of available network capacity.

+
+
+

Finally, HTTP/2 also enables more efficient processing of messages through use of binary message framing.

+
+
+
+

2. HTTP/2 Protocol Overview

+
+

HTTP/2 provides an optimized transport for HTTP semantics. HTTP/2 supports all of the core features of HTTP/1.1, but aims to be more efficient in several ways.

+
+
+

The basic protocol unit in HTTP/2 is a frame (Section 4.1). Each frame type serves a different purpose. For example, HEADERS and DATA frames form the basis of HTTP requests and responses (Section 8.1); other frame types like SETTINGS, WINDOW_UPDATE, and PUSH_PROMISE are used in support of other HTTP/2 features.

+
+
+

Multiplexing of requests is achieved by having each HTTP request-response exchange associated with its own stream (Section 5). Streams are largely independent of each other, so a blocked or stalled request or response does not prevent progress on other streams.

+
+
+

Flow control and prioritization ensure that it is possible to efficiently use multiplexed streams. Flow control (Section 5.2) helps to ensure that only data that can be used by a receiver is transmitted. Prioritization (Section 5.3) ensures that limited resources can be directed to the most important streams first.

+
+
+

HTTP/2 adds a new interaction mode, whereby a server can push responses to a client (Section 8.2). Server push allows a server to speculatively send data to a client that the server anticipates the client will need, trading off some network usage against a potential latency gain. The server does this by synthesizing a request, which it sends as a PUSH_PROMISE frame. The server is then able to send a response to the synthetic request on a separate stream.

+
+
+

Because HTTP header fields used in a connection can contain large amounts of redundant data, frames that contain them are compressed (Section 4.3). This has especially advantageous impact upon request sizes in the common case, allowing many requests to be compressed into one packet.

+
+
+

2.1 Document Organization +

+
+

The HTTP/2 specification is split into four parts:

+
    +
  • Starting HTTP/2 (Section 3) covers how an HTTP/2 connection is initiated.
  • +
  • The framing (Section 4) and streams (Section 5) layers describe the way HTTP/2 frames are structured and formed into multiplexed streams.
  • +
  • Frame (Section 6) and error (Section 7) definitions include details of the frame and error types used in HTTP/2.
  • +
  • HTTP mappings (Section 8) and additional requirements (Section 9) describe how HTTP semantics are expressed using frames and streams.
  • +
+
+
+

While some of the frame and stream layer concepts are isolated from HTTP, this specification does not define a completely generic framing layer. The framing and streams layers are tailored to the needs of the HTTP protocol and server push.

+
+
+
+

2.2 Conventions and Terminology +

+
+

The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in RFC 2119 + [RFC2119].

+
+
+

All numeric values are in network byte order. Values are unsigned unless otherwise indicated. Literal values are provided in decimal or hexadecimal as appropriate. Hexadecimal literals are prefixed with + 0x to distinguish them from decimal literals.

+
+
+

The following terms are used:

+
+
client:
+
The endpoint that initiates an HTTP/2 connection. Clients send HTTP requests and receive HTTP responses.
+
connection:
+
A transport-layer connection between two endpoints.
+
connection error:
+
An error that affects the entire HTTP/2 connection.
+
endpoint:
+
Either the client or server of the connection.
+
frame:
+
The smallest unit of communication within an HTTP/2 connection, consisting of a header and a variable-length sequence of octets structured according to the frame type.
+
peer:
+
An endpoint. When discussing a particular endpoint, "peer" refers to the endpoint that is remote to the primary subject of discussion.
+
receiver:
+
An endpoint that is receiving frames.
+
sender:
+
An endpoint that is transmitting frames.
+
server:
+
The endpoint that accepts an HTTP/2 connection. Servers receive HTTP requests and serve HTTP responses.
+
stream:
+
A bi-directional flow of frames within the HTTP/2 connection.
+
stream error:
+
An error on the individual HTTP/2 stream.
+
+
+
+

Finally, the terms "gateway", "intermediary", "proxy", and "tunnel" are defined in Section 2.3 of + + [RFC7230] + . Intermediaries act as both client and server at different times.

+
+
+

The term "payload body" is defined in Section 3.3 of + + [RFC7230] + .

+
+
+
+
+

3. Starting HTTP/2

+
+

An HTTP/2 connection is an application layer protocol running on top of a TCP connection ( + + [TCP] + ). The client is the TCP connection initiator.

+
+
+

HTTP/2 uses the same "http" and "https" URI schemes used by HTTP/1.1. HTTP/2 shares the same default port numbers: 80 for "http" URIs and 443 for "https" URIs. As a result, implementations processing requests for target resource URIs like + http://example.org/foo or + https://example.com/bar are required to first discover whether the upstream server (the immediate peer to which the client wishes to establish a connection) supports HTTP/2.

+
+
+

The means by which support for HTTP/2 is determined is different for "http" and "https" URIs. Discovery for "http" URIs is described in Section 3.2. Discovery for "https" URIs is described in Section 3.3.

+
+
+

3.1 HTTP/2 Version Identification

+
+

The protocol defined in this document has two identifiers.

+
    +
  • +

    The string "h2" identifies the protocol where HTTP/2 uses TLS + [TLS12]. This identifier is used in the TLS application layer protocol negotiation extension (ALPN) + [TLS-ALPN] field and in any place where HTTP/2 over TLS is identified.

    +

    The "h2" string is serialized into an ALPN protocol identifier as the two octet sequence: 0x68, 0x32.

    +
  • +
  • +

    The string "h2c" identifies the protocol where HTTP/2 is run over cleartext TCP. This identifier is used in the HTTP/1.1 Upgrade header field and in any place where HTTP/2 over TCP is identified.

    +

    The "h2c" string is reserved from the ALPN identifier space, but describes a protocol that does not use TLS.

    +
  • +
+
+
+

Negotiating "h2" or "h2c" implies the use of the transport, security, framing and message semantics described in this document.

+
+
+

[rfc.comment.1: RFC Editor's Note: please remove the remainder of this section prior to the publication of a final version of this document.]

+
+
+

Only implementations of the final, published RFC can identify themselves as "h2" or "h2c". Until such an RFC exists, implementations MUST NOT identify themselves using these strings.

+
+
+

Implementations of draft versions of the protocol MUST add the string "-" and the corresponding draft number to the identifier. For example, draft-ietf-httpbis-http2-11 over TLS is identified using the string "h2-11".

+
+
+

Non-compatible experiments that are based on these draft versions MUST append the string "-" and an experiment name to the identifier. For example, an experimental implementation of packet mood-based encoding based on draft-ietf-httpbis-http2-09 might identify itself as "h2-09-emo". Note that any label MUST conform to the "token" syntax defined in Section 3.2.6 of + + [RFC7230] + . Experimenters are encouraged to coordinate their experiments on the ietf-http-wg@w3.org mailing list.

+
+
+
+

3.2 Starting HTTP/2 for "http" URIs

+
+

A client that makes a request for an "http" URI without prior knowledge about support for HTTP/2 on the next hop uses the HTTP Upgrade mechanism (Section 6.7 of + + [RFC7230] + ). The client does so by making an HTTP/1.1 request that includes an Upgrade header field with the "h2c" token. Such an HTTP/1.1 request MUST include exactly one HTTP2-Settings (Section 3.2.1) header field.

+
+
+

For example:

GET / HTTP/1.1
+Host: server.example.com
+Connection: Upgrade, HTTP2-Settings
+Upgrade: h2c
+HTTP2-Settings: <base64url encoding of HTTP/2 SETTINGS payload>
+
+
+
+

Requests that contain an payload body MUST be sent in their entirety before the client can send HTTP/2 frames. This means that a large request can block the use of the connection until it is completely sent.

+
+
+

If concurrency of an initial request with subsequent requests is important, an OPTIONS request can be used to perform the upgrade to HTTP/2, at the cost of an additional round-trip.

+
+
+

A server that does not support HTTP/2 can respond to the request as though the Upgrade header field were absent:

+
+
HTTP/1.1 200 OK
+Content-Length: 243
+Content-Type: text/html
+
+...
+
+
+

A server MUST ignore an "h2" token in an Upgrade header field. Presence of a token with "h2" implies HTTP/2 over TLS, which is instead negotiated as described in Section 3.3.

+
+
+

A server that supports HTTP/2 accepts the upgrade with a 101 (Switching Protocols) response. After the empty line that terminates the 101 response, the server can begin sending HTTP/2 frames. These frames MUST include a response to the request that initiated the Upgrade.

+
+
+

For example:

HTTP/1.1 101 Switching Protocols
+Connection: Upgrade
+Upgrade: h2c
+
+[ HTTP/2 connection ...
+
+
+

The first HTTP/2 frame sent by the server MUST be a SETTINGS frame (Section 6.5) as the server connection preface (Section 3.5). Upon receiving the 101 response, the client MUST send a connection preface (Section 3.5), which includes a SETTINGS frame.

+
+
+

The HTTP/1.1 request that is sent prior to upgrade is assigned a stream identifier of 1 (see Section 5.1.1) with default priority values (Section 5.3.5). Stream 1 is implicitly "half closed" from the client toward the server (see Section 5.1), since the request is completed as an HTTP/1.1 request. After commencing the HTTP/2 connection, stream 1 is used for the response.

+
+
+

3.2.1 HTTP2-Settings Header Field

+
+

A request that upgrades from HTTP/1.1 to HTTP/2 MUST include exactly one + HTTP2-Settings header field. The + HTTP2-Settings header field is a connection-specific header field that includes parameters that govern the HTTP/2 connection, provided in anticipation of the server accepting the request to upgrade.

+
+
HTTP2-Settings    = token68
+
+
+

A server MUST NOT upgrade the connection to HTTP/2 if this header field is not present, or if more than one is present. A server MUST NOT send this header field.

+
+
+

The content of the + HTTP2-Settings header field is the payload of a SETTINGS frame (Section 6.5), encoded as a base64url string (that is, the URL- and filename-safe Base64 encoding described in Section 5 of + + [RFC4648] + , with any trailing '=' characters omitted). The ABNF + [RFC5234] production for + token68 is defined in Section 2.1 of + + [RFC7235] + .

+
+
+

Since the upgrade is only intended to apply to the immediate connection, a client sending + HTTP2-Settings MUST also send + HTTP2-Settings as a connection option in the + Connection header field to prevent it from being forwarded (see Section 6.1 of + + [RFC7230] + ).

+
+
+

A server decodes and interprets these values as it would any other SETTINGS frame. Explicit acknowledgement of these settings (Section 6.5.3) is not necessary, since a 101 response serves as implicit acknowledgment. Providing these values in the Upgrade request gives a client an opportunity to provide parameters prior to receiving any frames from the server.

+
+
+
+
+

3.3 Starting HTTP/2 for "https" URIs

+
+

A client that makes a request to an "https" URI uses TLS + [TLS12] with the application layer protocol negotiation (ALPN) extension + [TLS-ALPN].

+
+
+

HTTP/2 over TLS uses the "h2" protocol identifier. The "h2c" protocol identifier MUST NOT be sent by a client or selected by a server; the "h2c" protocol identifier describes a protocol that does not use TLS.

+
+
+

Once TLS negotiation is complete, both the client and the server MUST send a connection preface (Section 3.5).

+
+
+
+

3.4 Starting HTTP/2 with Prior Knowledge

+
+

A client can learn that a particular server supports HTTP/2 by other means. For example, + + [ALT-SVC] + describes a mechanism for advertising this capability.

+
+
+

A client MUST send the connection preface (Section 3.5), and then MAY immediately send HTTP/2 frames to such a server; servers can identify these connections by the presence of the connection preface. This only affects the establishment of HTTP/2 connections over cleartext TCP; implementations that support HTTP/2 over TLS MUST use protocol negotiation in TLS + [TLS-ALPN].

+
+
+

Likewise, the server MUST send a connection preface (Section 3.5).

+
+
+

Without additional information, prior support for HTTP/2 is not a strong signal that a given server will support HTTP/2 for future connections. For example, it is possible for server configurations to change, for configurations to differ between instances in clustered servers, or for network conditions to change.

+
+
+
+

3.5 HTTP/2 Connection Preface

+
+

In HTTP/2, each endpoint is required to send a connection preface as a final confirmation of the protocol in use, and to establish the initial settings for the HTTP/2 connection. The client and server each send a different connection preface.

+
+
+

The client connection preface starts with a sequence of 24 octets, which in hex notation are:

+
+
0x505249202a20485454502f322e300d0a0d0a534d0d0a0d0a
+
+
+

(the string + PRI * HTTP/2.0\r\n\r\nSM\r\n\r\n). This sequence MUST be followed by a SETTINGS frame (Section 6.5), which MAY be empty. The client sends the client connection preface immediately upon receipt of a 101 Switching Protocols response (indicating a successful upgrade), or as the first application data octets of a TLS connection. If starting an HTTP/2 connection with prior knowledge of server support for the protocol, the client connection preface is sent upon connection establishment.

+
+
+
    +
  • The client connection preface is selected so that a large proportion of HTTP/1.1 or HTTP/1.0 servers and intermediaries do not attempt to process further frames. Note that this does not address the concerns raised in + + [TALKING] + .
  • +
+
+
+

The server connection preface consists of a potentially empty SETTINGS frame (Section 6.5) that MUST be the first frame the server sends in the HTTP/2 connection.

+
+
+

The SETTINGS frames received from a peer as part of the connection preface MUST be acknowledged (see Section 6.5.3) after sending the connection preface.

+
+
+

To avoid unnecessary latency, clients are permitted to send additional frames to the server immediately after sending the client connection preface, without waiting to receive the server connection preface. It is important to note, however, that the server connection preface SETTINGS frame might include parameters that necessarily alter how a client is expected to communicate with the server. Upon receiving the SETTINGS frame, the client is expected to honor any parameters established. In some configurations, it is possible for the server to transmit SETTINGS before the client sends additional frames, providing an opportunity to avoid this issue.

+
+
+

Clients and servers MUST treat an invalid connection preface as a connection error (Section 5.4.1) of type PROTOCOL_ERROR. A GOAWAY frame (Section 6.8) MAY be omitted in this case, since an invalid preface indicates that the peer is not using HTTP/2.

+
+
+
+
+

4. HTTP Frames

+
+

Once the HTTP/2 connection is established, endpoints can begin exchanging frames.

+
+
+

4.1 Frame Format

+
+

All frames begin with a fixed 9-octet header followed by a variable-length payload.

+
+
+
 +-----------------------------------------------+
+ |                 Length (24)                   |
+ +---------------+---------------+---------------+
+ |   Type (8)    |   Flags (8)   |
+ +-+-------------+---------------+-------------------------------+
+ |R|                 Stream Identifier (31)                      |
+ +=+=============================================================+
+ |                   Frame Payload (0...)                      ...
+ +---------------------------------------------------------------+
+
+

Figure 1: Frame Layout

+
+

The fields of the frame header are defined as:

+
+
Length:
+
+

The length of the frame payload expressed as an unsigned 24-bit integer. Values greater than 214 (16,384) MUST NOT be sent unless the receiver has set a larger value for SETTINGS_MAX_FRAME_SIZE.

+

The 9 octets of the frame header are not included in this value.

+
+
Type:
+
+

The 8-bit type of the frame. The frame type determines the format and semantics of the frame. Implementations MUST ignore and discard any frame that has a type that is unknown.

+
+
Flags:
+
+

An 8-bit field reserved for frame-type specific boolean flags.

+

Flags are assigned semantics specific to the indicated frame type. Flags that have no defined semantics for a particular frame type MUST be ignored, and MUST be left unset (0x0) when sending.

+
+
R:
+
+

A reserved 1-bit field. The semantics of this bit are undefined and the bit MUST remain unset (0x0) when sending and MUST be ignored when receiving.

+
+
Stream Identifier:
+
+

A stream identifier (see Section 5.1.1) expressed as an unsigned 31-bit integer. The value 0x0 is reserved for frames that are associated with the connection as a whole as opposed to an individual stream.

+
+
+
+
+

The structure and content of the frame payload is dependent entirely on the frame type.

+
+
+
+

4.2 Frame Size

+
+

The size of a frame payload is limited by the maximum size that a receiver advertises in the SETTINGS_MAX_FRAME_SIZE setting. This setting can have any value between 214 (16,384) and 224-1 (16,777,215) octets, inclusive.

+
+
+

All implementations MUST be capable of receiving and minimally processing frames up to 214 octets in length, plus the 9 octet frame header (Section 4.1). The size of the frame header is not included when describing frame sizes.

+
+
Note:
+
Certain frame types, such as PING (Section 6.7), impose additional limits on the amount of payload data allowed.
+
+
+
+

An endpoint MUST send a FRAME_SIZE_ERROR error if a frame exceeds the size defined in SETTINGS_MAX_FRAME_SIZE, any limit defined for the frame type, or it is too small to contain mandatory frame data. A frame size error in a frame that could alter the state of the entire connection MUST be treated as a connection error (Section 5.4.1); this includes any frame carrying a header block (Section 4.3) (that is, HEADERS, PUSH_PROMISE, and CONTINUATION), SETTINGS, and any frame with a stream identifier of 0.

+
+
+

Endpoints are not obligated to use all available space in a frame. Responsiveness can be improved by using frames that are smaller than the permitted maximum size. Sending large frames can result in delays in sending time-sensitive frames (such as RST_STREAM, WINDOW_UPDATE, or PRIORITY) which if blocked by the transmission of a large frame, could affect performance.

+
+
+
+

4.3 Header Compression and Decompression

+
+

Just as in HTTP/1, a header field in HTTP/2 is a name with one or more associated values. They are used within HTTP request and response messages as well as server push operations (see Section 8.2).

+
+
+

Header lists are collections of zero or more header fields. When transmitted over a connection, a header list is serialized into a header block using HTTP Header Compression + [COMPRESSION]. The serialized header block is then divided into one or more octet sequences, called header block fragments, and transmitted within the payload of HEADERS (Section 6.2), PUSH_PROMISE (Section 6.6) or CONTINUATION (Section 6.10) frames.

+
+
+

The Cookie header field + [COOKIE] is treated specially by the HTTP mapping (see Section 8.1.2.5).

+
+
+

A receiving endpoint reassembles the header block by concatenating its fragments, then decompresses the block to reconstruct the header list.

+
+
+

A complete header block consists of either:

+ +
+
+

Header compression is stateful. One compression context and one decompression context is used for the entire connection. A decoding error in a header block MUST be treated as a connection error (Section 5.4.1) of type COMPRESSION_ERROR.

+
+
+

Each header block is processed as a discrete unit. Header blocks MUST be transmitted as a contiguous sequence of frames, with no interleaved frames of any other type or from any other stream. The last frame in a sequence of HEADERS or CONTINUATION frames has the END_HEADERS flag set. The last frame in a sequence of PUSH_PROMISE or CONTINUATION frames has the END_HEADERS flag set. This allows a header block to be logically equivalent to a single frame.

+
+
+

Header block fragments can only be sent as the payload of HEADERS, PUSH_PROMISE or CONTINUATION frames, because these frames carry data that can modify the compression context maintained by a receiver. An endpoint receiving HEADERS, PUSH_PROMISE or CONTINUATION frames needs to reassemble header blocks and perform decompression even if the frames are to be discarded. A receiver MUST terminate the connection with a connection error (Section 5.4.1) of type COMPRESSION_ERROR if it does not decompress a header block.

+
+
+
+
+

5. Streams and Multiplexing

+
+

A "stream" is an independent, bi-directional sequence of frames exchanged between the client and server within an HTTP/2 connection. Streams have several important characteristics:

+ +
+
+

5.1 Stream States

+
+

The lifecycle of a stream is shown in Figure 2.

+
+
+

Figure 2: Stream States

+
+

Note that this diagram shows stream state transitions and the frames and flags that affect those transitions only. In this regard, CONTINUATION frames do not result in state transitions; they are effectively part of the HEADERS or PUSH_PROMISE that they follow. For the purpose of state transitions, the END_STREAM flag is processed as a separate event to the frame that bears it; a HEADERS frame with the END_STREAM flag set can cause two state transitions.

+
+
+

Both endpoints have a subjective view of the state of a stream that could be different when frames are in transit. Endpoints do not coordinate the creation of streams; they are created unilaterally by either endpoint. The negative consequences of a mismatch in states are limited to the "closed" state after sending RST_STREAM, where frames might be received for some time after closing.

+
+
+

Streams have the following states:

+
+
idle:
+
+

+
All streams start in the "idle" state.

+

The following transitions are valid from this state:

+
    +
  • Sending or receiving a HEADERS frame causes the stream to become "open". The stream identifier is selected as described in Section 5.1.1. The same HEADERS frame can also cause a stream to immediately become "half closed".
  • +
  • Sending a PUSH_PROMISE frame on another stream reserves the idle stream that is identified for later use. The stream state for the reserved stream transitions to "reserved (local)".
  • +
  • Receiving a PUSH_PROMISE frame on another stream reserves an idle stream that is identified for later use. The stream state for the reserved stream transitions to "reserved (remote)".
  • +
  • Note that the PUSH_PROMISE frame is not sent on the idle stream, but references the newly reserved stream in the Promised Stream ID field.
  • +
+

Receiving any frame other than HEADERS or PRIORITY on a stream in this state MUST be treated as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
+
reserved (local):
+
+

+
A stream in the "reserved (local)" state is one that has been promised by sending a PUSH_PROMISE frame. A PUSH_PROMISE frame reserves an idle stream by associating the stream with an open stream that was initiated by the remote peer (see Section 8.2).

+

In this state, only the following transitions are possible:

+
    +
  • The endpoint can send a HEADERS frame. This causes the stream to open in a "half closed (remote)" state.
  • +
  • Either endpoint can send a RST_STREAM frame to cause the stream to become "closed". This releases the stream reservation.
  • +
+

An endpoint MUST NOT send any type of frame other than HEADERS, RST_STREAM, or PRIORITY in this state.

+

A PRIORITY or WINDOW_UPDATE frame MAY be received in this state. Receiving any type of frame other than RST_STREAM, PRIORITY or WINDOW_UPDATE on a stream in this state MUST be treated as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
+
reserved (remote):
+
+

+
A stream in the "reserved (remote)" state has been reserved by a remote peer.

+

In this state, only the following transitions are possible:

+
    +
  • Receiving a HEADERS frame causes the stream to transition to "half closed (local)".
  • +
  • Either endpoint can send a RST_STREAM frame to cause the stream to become "closed". This releases the stream reservation.
  • +
+

An endpoint MAY send a PRIORITY frame in this state to reprioritize the reserved stream. An endpoint MUST NOT send any type of frame other than RST_STREAM, WINDOW_UPDATE, or PRIORITY in this state.

+

Receiving any type of frame other than HEADERS, RST_STREAM or PRIORITY on a stream in this state MUST be treated as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
+
open:
+
+

+
A stream in the "open" state may be used by both peers to send frames of any type. In this state, sending peers observe advertised stream level flow control limits (Section 5.2).

+

From this state either endpoint can send a frame with an END_STREAM flag set, which causes the stream to transition into one of the "half closed" states: an endpoint sending an END_STREAM flag causes the stream state to become "half closed (local)"; an endpoint receiving an END_STREAM flag causes the stream state to become "half closed (remote)".

+

Either endpoint can send a RST_STREAM frame from this state, causing it to transition immediately to "closed".

+
+
half closed (local):
+
+

+
A stream that is in the "half closed (local)" state cannot be used for sending frames other than WINDOW_UPDATE, PRIORITY and RST_STREAM.

+

A stream transitions from this state to "closed" when a frame that contains an END_STREAM flag is received, or when either peer sends a RST_STREAM frame.

+

An endpoint can receive any type of frame in this state. Providing flow control credit using WINDOW_UPDATE frames is necessary to continue receiving flow controlled frames. A receiver can ignore WINDOW_UPDATE frames in this state, which might arrive for a short period after a frame bearing the END_STREAM flag is sent.

+

PRIORITY frames received in this state are used to reprioritize streams that depend on the identified stream.

+
+
half closed (remote):
+
+

+
A stream that is "half closed (remote)" is no longer being used by the peer to send frames. In this state, an endpoint is no longer obligated to maintain a receiver flow control window.

+

If an endpoint receives additional frames for a stream that is in this state, other than WINDOW_UPDATE, PRIORITY or RST_STREAM, it MUST respond with a stream error (Section 5.4.2) of type STREAM_CLOSED.

+

A stream that is "half closed (remote)" can be used by the endpoint to send frames of any type. In this state, the endpoint continues to observe advertised stream level flow control limits (Section 5.2).

+

A stream can transition from this state to "closed" by sending a frame that contains an END_STREAM flag, or when either peer sends a RST_STREAM frame.

+
+
closed:
+
+

+
The "closed" state is the terminal state.

+

An endpoint MUST NOT send frames other than PRIORITY on a closed stream. An endpoint that receives any frame other than PRIORITY after receiving a RST_STREAM MUST treat that as a stream error (Section 5.4.2) of type STREAM_CLOSED. Similarly, an endpoint that receives any frames after receiving a frame with the END_STREAM flag set MUST treat that as a connection error (Section 5.4.1) of type STREAM_CLOSED, unless the frame is permitted as described below.

+

WINDOW_UPDATE or RST_STREAM frames can be received in this state for a short period after a DATA or HEADERS frame containing an END_STREAM flag is sent. Until the remote peer receives and processes RST_STREAM or the frame bearing the END_STREAM flag, it might send frames of these types. Endpoints MUST ignore WINDOW_UPDATE or RST_STREAM frames received in this state, though endpoints MAY choose to treat frames that arrive a significant time after sending END_STREAM as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+

PRIORITY frames can be sent on closed streams to prioritize streams that are dependent on the closed stream. Endpoints SHOULD process PRIORITY frames, though they can be ignored if the stream has been removed from the dependency tree (see Section 5.3.4).

+

If this state is reached as a result of sending a RST_STREAM frame, the peer that receives the RST_STREAM might have already sent - or enqueued for sending - frames on the stream that cannot be withdrawn. An endpoint MUST ignore frames that it receives on closed streams after it has sent a RST_STREAM frame. An endpoint MAY choose to limit the period over which it ignores frames and treat frames that arrive after this time as being in error.

+

Flow controlled frames (i.e., DATA) received after sending RST_STREAM are counted toward the connection flow control window. Even though these frames might be ignored, because they are sent before the sender receives the RST_STREAM, the sender will consider the frames to count against the flow control window.

+

An endpoint might receive a PUSH_PROMISE frame after it sends RST_STREAM. PUSH_PROMISE causes a stream to become "reserved" even if the associated stream has been reset. Therefore, a RST_STREAM is needed to close an unwanted promised stream.

+
+
+
+
+

In the absence of more specific guidance elsewhere in this document, implementations SHOULD treat the receipt of a frame that is not expressly permitted in the description of a state as a connection error (Section 5.4.1) of type PROTOCOL_ERROR. Note that PRIORITY can be sent and received in any stream state. Frames of unknown types are ignored.

+
+
+

An example of the state transitions for an HTTP request/response exchange can be found in Section 8.1. An example of the state transitions for server push can be found in Section 8.2.1 and Section 8.2.2.

+
+
+

5.1.1 Stream Identifiers

+
+

Streams are identified with an unsigned 31-bit integer. Streams initiated by a client MUST use odd-numbered stream identifiers; those initiated by the server MUST use even-numbered stream identifiers. A stream identifier of zero (0x0) is used for connection control messages; the stream identifier zero cannot be used to establish a new stream.

+
+
+

HTTP/1.1 requests that are upgraded to HTTP/2 (see Section 3.2) are responded to with a stream identifier of one (0x1). After the upgrade completes, stream 0x1 is "half closed (local)" to the client. Therefore, stream 0x1 cannot be selected as a new stream identifier by a client that upgrades from HTTP/1.1.

+
+
+

The identifier of a newly established stream MUST be numerically greater than all streams that the initiating endpoint has opened or reserved. This governs streams that are opened using a HEADERS frame and streams that are reserved using PUSH_PROMISE. An endpoint that receives an unexpected stream identifier MUST respond with a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
+
+

The first use of a new stream identifier implicitly closes all streams in the "idle" state that might have been initiated by that peer with a lower-valued stream identifier. For example, if a client sends a HEADERS frame on stream 7 without ever sending a frame on stream 5, then stream 5 transitions to the "closed" state when the first frame for stream 7 is sent or received.

+
+
+

Stream identifiers cannot be reused. Long-lived connections can result in an endpoint exhausting the available range of stream identifiers. A client that is unable to establish a new stream identifier can establish a new connection for new streams. A server that is unable to establish a new stream identifier can send a GOAWAY frame so that the client is forced to open a new connection for new streams.

+
+
+
+

5.1.2 Stream Concurrency +

+
+

A peer can limit the number of concurrently active streams using the SETTINGS_MAX_CONCURRENT_STREAMS parameter (see Section 6.5.2) within a SETTINGS frame. The maximum concurrent streams setting is specific to each endpoint and applies only to the peer that receives the setting. That is, clients specify the maximum number of concurrent streams the server can initiate, and servers specify the maximum number of concurrent streams the client can initiate.

+
+
+

Streams that are in the "open" state, or either of the "half closed" states count toward the maximum number of streams that an endpoint is permitted to open. Streams in any of these three states count toward the limit advertised in the SETTINGS_MAX_CONCURRENT_STREAMS setting. Streams in either of the "reserved" states do not count toward the stream limit.

+
+
+

Endpoints MUST NOT exceed the limit set by their peer. An endpoint that receives a HEADERS frame that causes their advertised concurrent stream limit to be exceeded MUST treat this as a stream error (Section 5.4.2) of type PROTOCOL_ERROR or REFUSED_STREAM. The choice of error code determines whether the endpoint wishes to enable automatic retry, see Section 8.1.4) for details.

+
+
+

An endpoint that wishes to reduce the value of SETTINGS_MAX_CONCURRENT_STREAMS to a value that is below the current number of open streams can either close streams that exceed the new value or allow streams to complete.

+
+
+
+
+

5.2 Flow Control

+
+

Using streams for multiplexing introduces contention over use of the TCP connection, resulting in blocked streams. A flow control scheme ensures that streams on the same connection do not destructively interfere with each other. Flow control is used for both individual streams and for the connection as a whole.

+
+
+

HTTP/2 provides for flow control through use of the WINDOW_UPDATE frame (Section 6.9).

+
+
+

5.2.1 Flow Control Principles

+
+

HTTP/2 stream flow control aims to allow a variety of flow control algorithms to be used without requiring protocol changes. Flow control in HTTP/2 has the following characteristics:

+
    +
  1. Flow control is specific to a connection. Both types of flow control are between the endpoints of a single hop, and not over the entire end-to-end path.
  2. +
  3. Flow control is based on window update frames. Receivers advertise how many octets they are prepared to receive on a stream and for the entire connection. This is a credit-based scheme.
  4. +
  5. Flow control is directional with overall control provided by the receiver. A receiver MAY choose to set any window size that it desires for each stream and for the entire connection. A sender MUST respect flow control limits imposed by a receiver. Clients, servers and intermediaries all independently advertise their flow control window as a receiver and abide by the flow control limits set by their peer when sending.
  6. +
  7. The initial value for the flow control window is 65,535 octets for both new streams and the overall connection.
  8. +
  9. The frame type determines whether flow control applies to a frame. Of the frames specified in this document, only DATA frames are subject to flow control; all other frame types do not consume space in the advertised flow control window. This ensures that important control frames are not blocked by flow control.
  10. +
  11. Flow control cannot be disabled.
  12. +
  13. HTTP/2 defines only the format and semantics of the WINDOW_UPDATE frame (Section 6.9). This document does not stipulate how a receiver decides when to send this frame or the value that it sends, nor does it specify how a sender chooses to send packets. Implementations are able to select any algorithm that suits their needs.
  14. +
+
+
+

Implementations are also responsible for managing how requests and responses are sent based on priority; choosing how to avoid head of line blocking for requests; and managing the creation of new streams. Algorithm choices for these could interact with any flow control algorithm.

+
+
+
+

5.2.2 Appropriate Use of Flow Control

+
+

Flow control is defined to protect endpoints that are operating under resource constraints. For example, a proxy needs to share memory between many connections, and also might have a slow upstream connection and a fast downstream one. Flow control addresses cases where the receiver is unable to process data on one stream, yet wants to continue to process other streams in the same connection.

+
+
+

Deployments that do not require this capability can advertise a flow control window of the maximum size (231-1), and by maintaining this window by sending a WINDOW_UPDATE frame when any data is received. This effectively disables flow control for that receiver. Conversely, a sender is always subject to the flow control window advertised by the receiver.

+
+
+

Deployments with constrained resources (for example, memory) can employ flow control to limit the amount of memory a peer can consume. Note, however, that this can lead to suboptimal use of available network resources if flow control is enabled without knowledge of the bandwidth-delay product (see + + [RFC7323] + ).

+
+
+

Even with full awareness of the current bandwidth-delay product, implementation of flow control can be difficult. When using flow control, the receiver MUST read from the TCP receive buffer in a timely fashion. Failure to do so could lead to a deadlock when critical frames, such as WINDOW_UPDATE, are not read and acted upon.

+
+
+
+
+

5.3 Stream priority

+
+

A client can assign a priority for a new stream by including prioritization information in the HEADERS frame (Section 6.2) that opens the stream. At any other time, the PRIORITY frame (Section 6.3) can be used to change the priority of a stream.

+
+
+

The purpose of prioritization is to allow an endpoint to express how it would prefer its peer allocate resources when managing concurrent streams. Most importantly, priority can be used to select streams for transmitting frames when there is limited capacity for sending.

+
+
+

Streams can be prioritized by marking them as dependent on the completion of other streams (Section 5.3.1). Each dependency is assigned a relative weight, a number that is used to determine the relative proportion of available resources that are assigned to streams dependent on the same stream.

+
+
+

Explicitly setting the priority for a stream is input to a prioritization process. It does not guarantee any particular processing or transmission order for the stream relative to any other stream. An endpoint cannot force a peer to process concurrent streams in a particular order using priority. Expressing priority is therefore only ever a suggestion.

+
+
+

Prioritization information can be omitted from messages. Defaults are used prior to any explicit values being provided (Section 5.3.5).

+
+
+

5.3.1 Stream Dependencies

+
+

Each stream can be given an explicit dependency on another stream. Including a dependency expresses a preference to allocate resources to the identified stream rather than to the dependent stream.

+
+
+

A stream that is not dependent on any other stream is given a stream dependency of 0x0. In other words, the non-existent stream 0 forms the root of the tree.

+
+
+

A stream that depends on another stream is a dependent stream. The stream upon which a stream is dependent is a parent stream. A dependency on a stream that is not currently in the tree - such as a stream in the "idle" state - results in that stream being given a default priority (Section 5.3.5).

+
+
+

When assigning a dependency on another stream, the stream is added as a new dependency of the parent stream. Dependent streams that share the same parent are not ordered with respect to each other. For example, if streams B and C are dependent on stream A, and if stream D is created with a dependency on stream A, this results in a dependency order of A followed by B, C, and D in any order.

+
+
+
    A                 A
+   / \      ==>      /|\
+  B   C             B D C
+
+

Figure 3: Example of Default Dependency Creation

+
+

An exclusive flag allows for the insertion of a new level of dependencies. The exclusive flag causes the stream to become the sole dependency of its parent stream, causing other dependencies to become dependent on the exclusive stream. In the previous example, if stream D is created with an exclusive dependency on stream A, this results in D becoming the dependency parent of B and C.

+
+
+
                      A
+    A                 |
+   / \      ==>       D
+  B   C              / \
+                    B   C
+
+

Figure 4: Example of Exclusive Dependency Creation

+
+

Inside the dependency tree, a dependent stream SHOULD only be allocated resources if all of the streams that it depends on (the chain of parent streams up to 0x0) are either closed, or it is not possible to make progress on them.

+
+
+

A stream cannot depend on itself. An endpoint MUST treat this as a stream error (Section 5.4.2) of type PROTOCOL_ERROR.

+
+
+
+

5.3.2 Dependency Weighting +

+
+

All dependent streams are allocated an integer weight between 1 and 256 (inclusive).

+
+
+

Streams with the same parent SHOULD be allocated resources proportionally based on their weight. Thus, if stream B depends on stream A with weight 4, and C depends on stream A with weight 12, and if no progress can be made on A, stream B ideally receives one third of the resources allocated to stream C.

+
+
+
+

5.3.3 Reprioritization

+
+

Stream priorities are changed using the PRIORITY frame. Setting a dependency causes a stream to become dependent on the identified parent stream.

+
+
+

Dependent streams move with their parent stream if the parent is reprioritized. Setting a dependency with the exclusive flag for a reprioritized stream moves all the dependencies of the new parent stream to become dependent on the reprioritized stream.

+
+
+

If a stream is made dependent on one of its own dependencies, the formerly dependent stream is first moved to be dependent on the reprioritized stream's previous parent. The moved dependency retains its weight.

+
+
+
+

For example, consider an original dependency tree where B and C depend on A, D and E depend on C, and F depends on D. If A is made dependent on D, then D takes the place of A. All other dependency relationships stay the same, except for F, which becomes dependent on A if the reprioritization is exclusive.

    ?                ?                ?                 ?
+    |               / \               |                 |
+    A              D   A              D                 D
+   / \            /   / \            / \                |
+  B   C     ==>  F   B   C   ==>    F   A       OR      A
+     / \                 |             / \             /|\
+    D   E                E            B   C           B C F
+    |                                     |             |
+    F                                     E             E
+               (intermediate)   (non-exclusive)    (exclusive)
+
+

Figure 5: Example of Dependency Reordering

+
+
+

5.3.4 Prioritization State Management

+
+

When a stream is removed from the dependency tree, its dependencies can be moved to become dependent on the parent of the closed stream. The weights of new dependencies are recalculated by distributing the weight of the dependency of the closed stream proportionally based on the weights of its dependencies.

+
+
+

Streams that are removed from the dependency tree cause some prioritization information to be lost. Resources are shared between streams with the same parent stream, which means that if a stream in that set closes or becomes blocked, any spare capacity allocated to a stream is distributed to the immediate neighbors of the stream. However, if the common dependency is removed from the tree, those streams share resources with streams at the next highest level.

+
+
+

For example, assume streams A and B share a parent, and streams C and D both depend on stream A. Prior to the removal of stream A, if streams A and D are unable to proceed, then stream C receives all the resources dedicated to stream A. If stream A is removed from the tree, the weight of stream A is divided between streams C and D. If stream D is still unable to proceed, this results in stream C receiving a reduced proportion of resources. For equal starting weights, C receives one third, rather than one half, of available resources.

+
+
+

It is possible for a stream to become closed while prioritization information that creates a dependency on that stream is in transit. If a stream identified in a dependency has no associated priority information, then the dependent stream is instead assigned a default priority (Section 5.3.5). This potentially creates suboptimal prioritization, since the stream could be given a priority that is different to what is intended.

+
+
+

To avoid these problems, an endpoint SHOULD retain stream prioritization state for a period after streams become closed. The longer state is retained, the lower the chance that streams are assigned incorrect or default priority values.

+
+
+

Similarly, streams that are in the "idle" state can be assigned priority or become a parent of other streams. This allows for the creation of a grouping node in the dependency tree, which enables more flexible expressions of priority. Idle streams begin with a default priority (Section 5.3.5).

+
+
+

The retention of priority information for streams that are not counted toward the limit set by SETTINGS_MAX_CONCURRENT_STREAMS could create a large state burden for an endpoint. Therefore the amount of prioritization state that is retained MAY be limited.

+
+
+

The amount of additional state an endpoint maintains for prioritization could be dependent on load; under high load, prioritization state can be discarded to limit resource commitments. In extreme cases, an endpoint could even discard prioritization state for active or reserved streams. If a limit is applied, endpoints SHOULD maintain state for at least as many streams as allowed by their setting for SETTINGS_MAX_CONCURRENT_STREAMS. Implementations SHOULD also attempt to retain state for streams that are in active use in the priority tree.

+
+
+

An endpoint receiving a PRIORITY frame that changes the priority of a closed stream SHOULD alter the dependencies of the streams that depend on it, if it has retained enough state to do so.

+
+
+
+

5.3.5 Default Priorities

+
+

All streams are initially assigned a non-exclusive dependency on stream 0x0. Pushed streams (Section 8.2) initially depend on their associated stream. In both cases, streams are assigned a default weight of 16.

+
+
+
+
+

5.4 Error Handling

+
+

HTTP/2 framing permits two classes of error:

+
    +
  • An error condition that renders the entire connection unusable is a connection error.
  • +
  • An error in an individual stream is a stream error.
  • +
+
+
+

A list of error codes is included in Section 7.

+
+
+

5.4.1 Connection Error Handling

+
+

A connection error is any error which prevents further processing of the framing layer, or which corrupts any connection state.

+
+
+

An endpoint that encounters a connection error SHOULD first send a GOAWAY frame (Section 6.8) with the stream identifier of the last stream that it successfully received from its peer. The GOAWAY frame includes an error code that indicates why the connection is terminating. After sending the GOAWAY frame for an error condition, the endpoint MUST close the TCP connection.

+
+
+

It is possible that the GOAWAY will not be reliably received by the receiving endpoint (see + + [RFC7230] + , Section 6.6). In the event of a connection error, GOAWAY only provides a best effort attempt to communicate with the peer about why the connection is being terminated.

+
+
+

An endpoint can end a connection at any time. In particular, an endpoint MAY choose to treat a stream error as a connection error. Endpoints SHOULD send a GOAWAY frame when ending a connection, providing that circumstances permit it.

+
+
+
+

5.4.2 Stream Error Handling

+
+

A stream error is an error related to a specific stream that does not affect processing of other streams.

+
+
+

An endpoint that detects a stream error sends a RST_STREAM frame (Section 6.4) that contains the stream identifier of the stream where the error occurred. The RST_STREAM frame includes an error code that indicates the type of error.

+
+
+

A RST_STREAM is the last frame that an endpoint can send on a stream. The peer that sends the RST_STREAM frame MUST be prepared to receive any frames that were sent or enqueued for sending by the remote peer. These frames can be ignored, except where they modify connection state (such as the state maintained for header compression (Section 4.3), or flow control).

+
+
+

Normally, an endpoint SHOULD NOT send more than one RST_STREAM frame for any stream. However, an endpoint MAY send additional RST_STREAM frames if it receives frames on a closed stream after more than a round-trip time. This behavior is permitted to deal with misbehaving implementations.

+
+
+

An endpoint MUST NOT send a RST_STREAM in response to a RST_STREAM frame, to avoid looping.

+
+
+
+

5.4.3 Connection Termination +

+
+

If the TCP connection is closed or reset while streams remain in open or half closed states, then the affected streams cannot be automatically retried (see Section 8.1.4 for details).

+
+
+
+
+

5.5 Extending HTTP/2

+
+

HTTP/2 permits extension of the protocol. Protocol extensions can be used to provide additional services or alter any aspect of the protocol, within the limitations described in this section. Extensions are effective only within the scope of a single HTTP/2 connection.

+
+
+

This applies to the protocol elements defined in this document. This does not affect the existing options for extending HTTP, such as defining new methods, status codes, or header fields.

+
+
+

Extensions are permitted to use new frame types (Section 4.1), new settings (Section 6.5.2), or new error codes (Section 7). Registries are established for managing these extension points: frame types (Section 11.2), settings (Section 11.3) and error codes (Section 11.4).

+
+
+

Implementations MUST ignore unknown or unsupported values in all extensible protocol elements. Implementations MUST discard frames that have unknown or unsupported types. This means that any of these extension points can be safely used by extensions without prior arrangement or negotiation. However, extension frames that appear in the middle of a header block (Section 4.3) are not permitted; these MUST be treated as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
+
+

Extensions that could change the semantics of existing protocol components MUST be negotiated before being used. For example, an extension that changes the layout of the HEADERS frame cannot be used until the peer has given a positive signal that this is acceptable. In this case, it could also be necessary to coordinate when the revised layout comes into effect. Note that treating any frame other than DATA frames as flow controlled is such a change in semantics, and can only be done through negotiation.

+
+
+

This document doesn't mandate a specific method for negotiating the use of an extension, but notes that a setting (Section 6.5.2) could be used for that purpose. If both peers set a value that indicates willingness to use the extension, then the extension can be used. If a setting is used for extension negotiation, the initial value MUST be defined in such a fashion that the extension is initially disabled.

+
+
+
+
+

6. Frame Definitions

+
+

This specification defines a number of frame types, each identified by a unique 8-bit type code. Each frame type serves a distinct purpose either in the establishment and management of the connection as a whole, or of individual streams.

+
+
+

The transmission of specific frame types can alter the state of a connection. If endpoints fail to maintain a synchronized view of the connection state, successful communication within the connection will no longer be possible. Therefore, it is important that endpoints have a shared comprehension of how the state is affected by the use any given frame.

+
+
+

6.1 DATA

+
+

DATA frames (type=0x0) convey arbitrary, variable-length sequences of octets associated with a stream. One or more DATA frames are used, for instance, to carry HTTP request or response payloads.

+
+
+

DATA frames MAY also contain padding. Padding can be added to DATA frames to obscure the size of messages. Padding is a security feature; see Section 10.7.

+
+
+
 +---------------+
+ |Pad Length? (8)|
+ +---------------+-----------------------------------------------+
+ |                            Data (*)                         ...
+ +---------------------------------------------------------------+
+ |                           Padding (*)                       ...
+ +---------------------------------------------------------------+
+
+

Figure 6: DATA Frame Payload

+
+

The DATA frame contains the following fields:

+
+
Pad Length:
+
An 8-bit field containing the length of the frame padding in units of octets. This field is conditional and is only present if the PADDED flag is set.
+
Data:
+
Application data. The amount of data is the remainder of the frame payload after subtracting the length of the other fields that are present.
+
Padding:
+
Padding octets that contain no application semantic value. Padding octets MUST be set to zero when sending. A receiver is not obligated to verify padding, but MAY treat non-zero padding as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.
+
+
+
+

The DATA frame defines the following flags:

+
+
END_STREAM (0x1):
+
Bit 0 being set indicates that this frame is the last that the endpoint will send for the identified stream. Setting this flag causes the stream to enter one of the "half closed" states or the "closed" state (Section 5.1).
+
PADDED (0x8):
+
Bit 3 being set indicates that the Pad Length field and any padding that it describes is present.
+
+
+
+

DATA frames MUST be associated with a stream. If a DATA frame is received whose stream identifier field is 0x0, the recipient MUST respond with a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
+
+

DATA frames are subject to flow control and can only be sent when a stream is in the "open" or "half closed (remote)" states. The entire DATA frame payload is included in flow control, including Pad Length and Padding fields if present. If a DATA frame is received whose stream is not in "open" or "half closed (local)" state, the recipient MUST respond with a stream error (Section 5.4.2) of type STREAM_CLOSED.

+
+
+

The total number of padding octets is determined by the value of the Pad Length field. If the length of the padding is the length of the frame payload or greater, the recipient MUST treat this as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
+
Note:
+
A frame can be increased in size by one octet by including a Pad Length field with a value of zero.
+
+
+
+
+

6.2 HEADERS

+
+

The HEADERS frame (type=0x1) is used to open a stream (Section 5.1), and additionally carries a header block fragment. HEADERS frames can be sent on a stream in the "open" or "half closed (remote)" states.

+
+
+
 +---------------+
+ |Pad Length? (8)|
+ +-+-------------+-----------------------------------------------+
+ |E|                 Stream Dependency? (31)                     |
+ +-+-------------+-----------------------------------------------+
+ |  Weight? (8)  |
+ +-+-------------+-----------------------------------------------+
+ |                   Header Block Fragment (*)                 ...
+ +---------------------------------------------------------------+
+ |                           Padding (*)                       ...
+ +---------------------------------------------------------------+
+
+

Figure 7: HEADERS Frame Payload

+
+

The HEADERS frame payload has the following fields:

+
+
Pad Length:
+
An 8-bit field containing the length of the frame padding in units of octets. This field is only present if the PADDED flag is set.
+
E:
+
A single bit flag indicates that the stream dependency is exclusive, see Section 5.3. This field is only present if the PRIORITY flag is set.
+
Stream Dependency:
+
A 31-bit stream identifier for the stream that this stream depends on, see Section 5.3. This field is only present if the PRIORITY flag is set.
+
Weight:
+
An unsigned 8-bit integer representing a priority weight for the stream, see Section 5.3. Add one to the value to obtain a weight between 1 and 256. This field is only present if the PRIORITY flag is set.
+
Header Block Fragment:
+
A header block fragment (Section 4.3).
+
Padding:
+
Padding octets.
+
+
+
+

The HEADERS frame defines the following flags:

+
+
END_STREAM (0x1):
+
+

Bit 0 being set indicates that the header block (Section 4.3) is the last that the endpoint will send for the identified stream.

+

A HEADERS frame carries the END_STREAM flag that signals the end of a stream. However, a HEADERS frame with the END_STREAM flag set can be followed by CONTINUATION frames on the same stream. Logically, the CONTINUATION frames are part of the HEADERS frame.

+
+
END_HEADERS (0x4):
+
+

Bit 2 being set indicates that this frame contains an entire header block (Section 4.3) and is not followed by any CONTINUATION frames.

+

A HEADERS frame without the END_HEADERS flag set MUST be followed by a CONTINUATION frame for the same stream. A receiver MUST treat the receipt of any other type of frame or a frame on a different stream as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
+
PADDED (0x8):
+
+

Bit 3 being set indicates that the Pad Length field and any padding that it describes is present.

+
+
PRIORITY (0x20):
+
+

Bit 5 being set indicates that the Exclusive Flag (E), Stream Dependency, and Weight fields are present; see Section 5.3.

+
+
+
+
+

The payload of a HEADERS frame contains a header block fragment (Section 4.3). A header block that does not fit within a HEADERS frame is continued in a CONTINUATION frame (Section 6.10).

+
+
+

HEADERS frames MUST be associated with a stream. If a HEADERS frame is received whose stream identifier field is 0x0, the recipient MUST respond with a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
+
+

The HEADERS frame changes the connection state as described in Section 4.3.

+
+
+

The HEADERS frame can include padding. Padding fields and flags are identical to those defined for DATA frames (Section 6.1).

+
+
+

Prioritization information in a HEADERS frame is logically equivalent to a separate PRIORITY frame, but inclusion in HEADERS avoids the potential for churn in stream prioritization when new streams are created. Prioritization fields in HEADERS frames subsequent to the first on a stream reprioritize the stream (Section 5.3.3).

+
+
+
+

6.3 PRIORITY

+
+

The PRIORITY frame (type=0x2) specifies the sender-advised priority of a stream (Section 5.3). It can be sent at any time for any stream, including idle or closed streams.

+
+
+
 +-+-------------------------------------------------------------+
+ |E|                  Stream Dependency (31)                     |
+ +-+-------------+-----------------------------------------------+
+ |   Weight (8)  |
+ +-+-------------+
+
+

Figure 8: PRIORITY Frame Payload

+
+

The payload of a PRIORITY frame contains the following fields:

+
+
E:
+
A single bit flag indicates that the stream dependency is exclusive, see Section 5.3.
+
Stream Dependency:
+
A 31-bit stream identifier for the stream that this stream depends on, see Section 5.3.
+
Weight:
+
An unsigned 8-bit integer representing a priority weight for the stream, see Section 5.3. Add one to the value to obtain a weight between 1 and 256.
+
+
+
+

The PRIORITY frame does not define any flags.

+
+
+

The PRIORITY frame always identifies a stream. If a PRIORITY frame is received with a stream identifier of 0x0, the recipient MUST respond with a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
+
+

The PRIORITY frame can be sent on a stream in any state, though it cannot be sent between consecutive frames that comprise a single header block (Section 4.3). Note that this frame could arrive after processing or frame sending has completed, which would cause it to have no effect on the identified stream. For a stream that is in the "half closed (remote)" or "closed" - state, this frame can only affect processing of the identified stream and its dependent streams and not frame transmission on that stream.

+
+
+

The PRIORITY frame can be sent for a stream in the "idle" or "closed" states. This allows for the reprioritization of a group of dependent streams by altering the priority of an unused or closed parent stream.

+
+
+

A PRIORITY frame with a length other than 5 octets MUST be treated as a stream error (Section 5.4.2) of type FRAME_SIZE_ERROR.

+
+
+
+

6.4 RST_STREAM

+
+

The RST_STREAM frame (type=0x3) allows for immediate termination of a stream. RST_STREAM is sent to request cancellation of a stream, or to indicate that an error condition has occurred.

+
+
+
 +---------------------------------------------------------------+
+ |                        Error Code (32)                        |
+ +---------------------------------------------------------------+
+
+

Figure 9: RST_STREAM Frame Payload

+
+

The RST_STREAM frame contains a single unsigned, 32-bit integer identifying the error code (Section 7). The error code indicates why the stream is being terminated.

+
+
+

The RST_STREAM frame does not define any flags.

+
+
+

The RST_STREAM frame fully terminates the referenced stream and causes it to enter the closed state. After receiving a RST_STREAM on a stream, the receiver MUST NOT send additional frames for that stream, with the exception of PRIORITY. However, after sending the RST_STREAM, the sending endpoint MUST be prepared to receive and process additional frames sent on the stream that might have been sent by the peer prior to the arrival of the RST_STREAM.

+
+
+

RST_STREAM frames MUST be associated with a stream. If a RST_STREAM frame is received with a stream identifier of 0x0, the recipient MUST treat this as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
+
+

RST_STREAM frames MUST NOT be sent for a stream in the "idle" state. If a RST_STREAM frame identifying an idle stream is received, the recipient MUST treat this as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
+
+

A RST_STREAM frame with a length other than 4 octets MUST be treated as a connection error (Section 5.4.1) of type FRAME_SIZE_ERROR.

+
+
+
+

6.5 SETTINGS

+
+

The SETTINGS frame (type=0x4) conveys configuration parameters that affect how endpoints communicate, such as preferences and constraints on peer behavior. The SETTINGS frame is also used to acknowledge the receipt of those parameters. Individually, a SETTINGS parameter can also be referred to as a "setting".

+
+
+

SETTINGS parameters are not negotiated; they describe characteristics of the sending peer, which are used by the receiving peer. Different values for the same parameter can be advertised by each peer. For example, a client might set a high initial flow control window, whereas a server might set a lower value to conserve resources.

+
+
+

A SETTINGS frame MUST be sent by both endpoints at the start of a connection, and MAY be sent at any other time by either endpoint over the lifetime of the connection. Implementations MUST support all of the parameters defined by this specification.

+
+
+

Each parameter in a SETTINGS frame replaces any existing value for that parameter. Parameters are processed in the order in which they appear, and a receiver of a SETTINGS frame does not need to maintain any state other than the current value of its parameters. Therefore, the value of a SETTINGS parameter is the last value that is seen by a receiver.

+
+
+

SETTINGS parameters are acknowledged by the receiving peer. To enable this, the SETTINGS frame defines the following flag:

+
+
ACK (0x1):
+
Bit 0 being set indicates that this frame acknowledges receipt and application of the peer's SETTINGS frame. When this bit is set, the payload of the SETTINGS frame MUST be empty. Receipt of a SETTINGS frame with the ACK flag set and a length field value other than 0 MUST be treated as a connection error (Section 5.4.1) of type FRAME_SIZE_ERROR. For more info, see Settings Synchronization (Section 6.5.3).
+
+
+
+

SETTINGS frames always apply to a connection, never a single stream. The stream identifier for a SETTINGS frame MUST be zero (0x0). If an endpoint receives a SETTINGS frame whose stream identifier field is anything other than 0x0, the endpoint MUST respond with a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
+
+

The SETTINGS frame affects connection state. A badly formed or incomplete SETTINGS frame MUST be treated as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
+
+

A SETTINGS frame with a length other than a multiple of 6 octets MUST be treated as a connection error (Section 5.4.1) of type FRAME_SIZE_ERROR.

+
+
+

6.5.1 SETTINGS Format

+
+

The payload of a SETTINGS frame consists of zero or more parameters, each consisting of an unsigned 16-bit setting identifier and an unsigned 32-bit value.

+
+
+
 +-------------------------------+
+ |       Identifier (16)         |
+ +-------------------------------+-------------------------------+
+ |                        Value (32)                             |
+ +---------------------------------------------------------------+
+
+

Figure 10: Setting Format

+
+
+

6.5.2 Defined SETTINGS Parameters

+
+

The following parameters are defined:

+
+
SETTINGS_HEADER_TABLE_SIZE (0x1):
+
+

Allows the sender to inform the remote endpoint of the maximum size of the header compression table used to decode header blocks, in octets. The encoder can select any size equal to or less than this value by using signaling specific to the header compression format inside a header block, see + + [COMPRESSION] + . The initial value is 4,096 octets.

+
+
SETTINGS_ENABLE_PUSH (0x2):
+
+

This setting can be use to disable server push (Section 8.2). An endpoint MUST NOT send a PUSH_PROMISE frame if it receives this parameter set to a value of 0. An endpoint that has both set this parameter to 0 and had it acknowledged MUST treat the receipt of a PUSH_PROMISE frame as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+

The initial value is 1, which indicates that server push is permitted. Any value other than 0 or 1 MUST be treated as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
+
SETTINGS_MAX_CONCURRENT_STREAMS (0x3):
+
+

Indicates the maximum number of concurrent streams that the sender will allow. This limit is directional: it applies to the number of streams that the sender permits the receiver to create. Initially there is no limit to this value. It is recommended that this value be no smaller than 100, so as to not unnecessarily limit parallelism.

+

A value of 0 for SETTINGS_MAX_CONCURRENT_STREAMS SHOULD NOT be treated as special by endpoints. A zero value does prevent the creation of new streams, however this can also happen for any limit that is exhausted with active streams. Servers SHOULD only set a zero value for short durations; if a server does not wish to accept requests, closing the connection is more appropriate.

+
+
SETTINGS_INITIAL_WINDOW_SIZE (0x4):
+
+

Indicates the sender's initial window size (in octets) for stream level flow control. The initial value is 216-1 (65,535) octets.

+

This setting affects the window size of all streams, see Section 6.9.2.

+

Values above the maximum flow control window size of 231-1 MUST be treated as a connection error (Section 5.4.1) of type FLOW_CONTROL_ERROR.

+
+
SETTINGS_MAX_FRAME_SIZE (0x5):
+
+

Indicates the size of the largest frame payload that the sender is willing to receive, in octets.

+

The initial value is 214 (16,384) octets. The value advertised by an endpoint MUST be between this initial value and the maximum allowed frame size (224-1 or 16,777,215 octets), inclusive. Values outside this range MUST be treated as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
+
SETTINGS_MAX_HEADER_LIST_SIZE (0x6):
+
+

This advisory setting informs a peer of the maximum size of header list that the sender is prepared to accept, in octets. The value is based on the uncompressed size of header fields, including the length of the name and value in octets plus an overhead of 32 octets for each header field.

+

For any given request, a lower limit than what is advertised MAY be enforced. The initial value of this setting is unlimited.

+
+
+
+
+

An endpoint that receives a SETTINGS frame with any unknown or unsupported identifier MUST ignore that setting.

+
+
+
+

6.5.3 Settings Synchronization

+
+

Most values in SETTINGS benefit from or require an understanding of when the peer has received and applied the changed parameter values. In order to provide such synchronization timepoints, the recipient of a SETTINGS frame in which the ACK flag is not set MUST apply the updated parameters as soon as possible upon receipt.

+
+
+

The values in the SETTINGS frame MUST be processed in the order they appear, with no other frame processing between values. Unsupported parameters MUST be ignored. Once all values have been processed, the recipient MUST immediately emit a SETTINGS frame with the ACK flag set. Upon receiving a SETTINGS frame with the ACK flag set, the sender of the altered parameters can rely on the setting having been applied.

+
+
+

If the sender of a SETTINGS frame does not receive an acknowledgement within a reasonable amount of time, it MAY issue a connection error (Section 5.4.1) of type SETTINGS_TIMEOUT.

+
+
+
+
+

6.6 PUSH_PROMISE

+
+

The PUSH_PROMISE frame (type=0x5) is used to notify the peer endpoint in advance of streams the sender intends to initiate. The PUSH_PROMISE frame includes the unsigned 31-bit identifier of the stream the endpoint plans to create along with a set of headers that provide additional context for the stream. Section 8.2 contains a thorough description of the use of PUSH_PROMISE frames.

+
+
+
 +---------------+
+ |Pad Length? (8)|
+ +-+-------------+-----------------------------------------------+
+ |R|                  Promised Stream ID (31)                    |
+ +-+-----------------------------+-------------------------------+
+ |                   Header Block Fragment (*)                 ...
+ +---------------------------------------------------------------+
+ |                           Padding (*)                       ...
+ +---------------------------------------------------------------+
+
+

Figure 11: PUSH_PROMISE Payload Format

+
+

The PUSH_PROMISE frame payload has the following fields:

+
+
Pad Length:
+
An 8-bit field containing the length of the frame padding in units of octets. This field is only present if the PADDED flag is set.
+
R:
+
A single reserved bit.
+
Promised Stream ID:
+
An unsigned 31-bit integer that identifies the stream that is reserved by the PUSH_PROMISE. The promised stream identifier MUST be a valid choice for the next stream sent by the sender (see new stream identifier (Section 5.1.1)).
+
Header Block Fragment:
+
A header block fragment (Section 4.3) containing request header fields.
+
Padding:
+
Padding octets.
+
+
+
+

The PUSH_PROMISE frame defines the following flags:

+
+
END_HEADERS (0x4):
+
+

Bit 2 being set indicates that this frame contains an entire header block (Section 4.3) and is not followed by any CONTINUATION frames.

+

A PUSH_PROMISE frame without the END_HEADERS flag set MUST be followed by a CONTINUATION frame for the same stream. A receiver MUST treat the receipt of any other type of frame or a frame on a different stream as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
+
PADDED (0x8):
+
+

Bit 3 being set indicates that the Pad Length field and any padding that it describes is present.

+
+
+
+
+

PUSH_PROMISE frames MUST only be sent on a peer-initiated stream that is in either the "open" or "half closed (remote)" state. The stream identifier of a PUSH_PROMISE frame indicates the stream it is associated with. If the stream identifier field specifies the value 0x0, a recipient MUST respond with a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
+
+

Promised streams are not required to be used in the order they are promised. The PUSH_PROMISE only reserves stream identifiers for later use.

+
+
+

PUSH_PROMISE MUST NOT be sent if the SETTINGS_ENABLE_PUSH setting of the peer endpoint is set to 0. An endpoint that has set this setting and has received acknowledgement MUST treat the receipt of a PUSH_PROMISE frame as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
+
+

Recipients of PUSH_PROMISE frames can choose to reject promised streams by returning a RST_STREAM referencing the promised stream identifier back to the sender of the PUSH_PROMISE.

+
+
+

A PUSH_PROMISE frame modifies the connection state in two ways. The inclusion of a header block (Section 4.3) potentially modifies the state maintained for header compression. PUSH_PROMISE also reserves a stream for later use, causing the promised stream to enter the "reserved" state. A sender MUST NOT send a PUSH_PROMISE on a stream unless that stream is either "open" or "half closed (remote)"; the sender MUST ensure that the promised stream is a valid choice for a new stream identifier (Section 5.1.1) (that is, the promised stream MUST be in the "idle" state).

+
+
+

Since PUSH_PROMISE reserves a stream, ignoring a PUSH_PROMISE frame causes the stream state to become indeterminate. A receiver MUST treat the receipt of a PUSH_PROMISE on a stream that is neither "open" nor "half closed (local)" as a connection error (Section 5.4.1) of type PROTOCOL_ERROR. However, an endpoint that has sent RST_STREAM on the associated stream MUST handle PUSH_PROMISE frames that might have been created before the RST_STREAM frame is received and processed.

+
+
+

A receiver MUST treat the receipt of a PUSH_PROMISE that promises an illegal stream identifier (Section 5.1.1) (that is, an identifier for a stream that is not currently in the "idle" state) as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
+
+

The PUSH_PROMISE frame can include padding. Padding fields and flags are identical to those defined for DATA frames (Section 6.1).

+
+
+
+

6.7 PING

+
+

The PING frame (type=0x6) is a mechanism for measuring a minimal round trip time from the sender, as well as determining whether an idle connection is still functional. PING frames can be sent from any endpoint.

+
+
+
 +---------------------------------------------------------------+
+ |                                                               |
+ |                      Opaque Data (64)                         |
+ |                                                               |
+ +---------------------------------------------------------------+
+
+

Figure 12: PING Payload Format

+
+

In addition to the frame header, PING frames MUST contain 8 octets of data in the payload. A sender can include any value it chooses and use those octets in any fashion.

+
+
+

Receivers of a PING frame that does not include an ACK flag MUST send a PING frame with the ACK flag set in response, with an identical payload. PING responses SHOULD be given higher priority than any other frame.

+
+
+

The PING frame defines the following flags:

+
+
ACK (0x1):
+
Bit 0 being set indicates that this PING frame is a PING response. An endpoint MUST set this flag in PING responses. An endpoint MUST NOT respond to PING frames containing this flag.
+
+
+
+

PING frames are not associated with any individual stream. If a PING frame is received with a stream identifier field value other than 0x0, the recipient MUST respond with a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
+
+

Receipt of a PING frame with a length field value other than 8 MUST be treated as a connection error (Section 5.4.1) of type FRAME_SIZE_ERROR.

+
+
+
+

6.8 GOAWAY

+
+

The GOAWAY frame (type=0x7) informs the remote peer to stop creating streams on this connection. GOAWAY can be sent by either the client or the server. Once sent, the sender will ignore frames sent on any new streams with identifiers higher than the included last stream identifier. Receivers of a GOAWAY frame MUST NOT open additional streams on the connection, although a new connection can be established for new streams.

+
+
+

The purpose of this frame is to allow an endpoint to gracefully stop accepting new streams, while still finishing processing of previously established streams. This enables administrative actions, like server maintenance.

+
+
+

There is an inherent race condition between an endpoint starting new streams and the remote sending a GOAWAY frame. To deal with this case, the GOAWAY contains the stream identifier of the last peer-initiated stream which was or might be processed on the sending endpoint in this connection. For instance, if the server sends a GOAWAY frame, the identified stream is the highest numbered stream initiated by the client.

+
+
+

If the receiver of the GOAWAY has sent data on streams with a higher stream identifier than what is indicated in the GOAWAY frame, those streams are not or will not be processed. The receiver of the GOAWAY frame can treat the streams as though they had never been created at all, thereby allowing those streams to be retried later on a new connection.

+
+
+

Endpoints SHOULD always send a GOAWAY frame before closing a connection so that the remote peer can know whether a stream has been partially processed or not. For example, if an HTTP client sends a POST at the same time that a server closes a connection, the client cannot know if the server started to process that POST request if the server does not send a GOAWAY frame to indicate what streams it might have acted on.

+
+
+

An endpoint might choose to close a connection without sending GOAWAY for misbehaving peers.

+
+
+
 +-+-------------------------------------------------------------+
+ |R|                  Last-Stream-ID (31)                        |
+ +-+-------------------------------------------------------------+
+ |                      Error Code (32)                          |
+ +---------------------------------------------------------------+
+ |                  Additional Debug Data (*)                    |
+ +---------------------------------------------------------------+
+
+

Figure 13: GOAWAY Payload Format

+
+

The GOAWAY frame does not define any flags.

+
+
+

The GOAWAY frame applies to the connection, not a specific stream. An endpoint MUST treat a GOAWAY frame with a stream identifier other than 0x0 as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
+
+

The last stream identifier in the GOAWAY frame contains the highest numbered stream identifier for which the sender of the GOAWAY frame might have taken some action on, or might yet take action on. All streams up to and including the identified stream might have been processed in some way. The last stream identifier can be set to 0 if no streams were processed.

+
+
Note:
+
In this context, "processed" means that some data from the stream was passed to some higher layer of software that might have taken some action as a result.
+
+

If a connection terminates without a GOAWAY frame, the last stream identifier is effectively the highest possible stream identifier.

+
+
+

On streams with lower or equal numbered identifiers that were not closed completely prior to the connection being closed, re-attempting requests, transactions, or any protocol activity is not possible, with the exception of idempotent actions like HTTP GET, PUT, or DELETE. Any protocol activity that uses higher numbered streams can be safely retried using a new connection.

+
+
+

Activity on streams numbered lower or equal to the last stream identifier might still complete successfully. The sender of a GOAWAY frame might gracefully shut down a connection by sending a GOAWAY frame, maintaining the connection in an open state until all in-progress streams complete.

+
+
+

An endpoint MAY send multiple GOAWAY frames if circumstances change. For instance, an endpoint that sends GOAWAY with NO_ERROR during graceful shutdown could subsequently encounter a condition that requires immediate termination of the connection. The last stream identifier from the last GOAWAY frame received indicates which streams could have been acted upon. Endpoints MUST NOT increase the value they send in the last stream identifier, since the peers might already have retried unprocessed requests on another connection.

+
+
+

A client that is unable to retry requests loses all requests that are in flight when the server closes the connection. This is especially true for intermediaries that might not be serving clients using HTTP/2. A server that is attempting to gracefully shut down a connection SHOULD send an initial GOAWAY frame with the last stream identifier set to 231-1 and a NO_ERROR code. This signals to the client that a shutdown is imminent and that no further requests can be initiated. After waiting at least one round trip time, the server can send another GOAWAY frame with an updated last stream identifier. This ensures that a connection can be cleanly shut down without losing requests.

+
+
+

After sending a GOAWAY frame, the sender can discard frames for streams with identifiers higher than the identified last stream. However, any frames that alter connection state cannot be completely ignored. For instance, HEADERS, PUSH_PROMISE and CONTINUATION frames MUST be minimally processed to ensure the state maintained for header compression is consistent (see Section 4.3); similarly DATA frames MUST be counted toward the connection flow control window. Failure to process these frames can cause flow control or header compression state to become unsynchronized.

+
+
+

The GOAWAY frame also contains a 32-bit error code (Section 7) that contains the reason for closing the connection.

+
+
+

Endpoints MAY append opaque data to the payload of any GOAWAY frame. Additional debug data is intended for diagnostic purposes only and carries no semantic value. Debug information could contain security- or privacy-sensitive data. Logged or otherwise persistently stored debug data MUST have adequate safeguards to prevent unauthorized access.

+
+
+
+

6.9 WINDOW_UPDATE

+
+

The WINDOW_UPDATE frame (type=0x8) is used to implement flow control; see Section 5.2 for an overview.

+
+
+

Flow control operates at two levels: on each individual stream and on the entire connection.

+
+
+

Both types of flow control are hop-by-hop; that is, only between the two endpoints. Intermediaries do not forward WINDOW_UPDATE frames between dependent connections. However, throttling of data transfer by any receiver can indirectly cause the propagation of flow control information toward the original sender.

+
+
+

Flow control only applies to frames that are identified as being subject to flow control. Of the frame types defined in this document, this includes only DATA frames. Frames that are exempt from flow control MUST be accepted and processed, unless the receiver is unable to assign resources to handling the frame. A receiver MAY respond with a stream error (Section 5.4.2) or connection error (Section 5.4.1) of type FLOW_CONTROL_ERROR if it is unable to accept a frame.

+
+
+
 +-+-------------------------------------------------------------+
+ |R|              Window Size Increment (31)                     |
+ +-+-------------------------------------------------------------+
+
+

Figure 14: WINDOW_UPDATE Payload Format

+
+

The payload of a WINDOW_UPDATE frame is one reserved bit, plus an unsigned 31-bit integer indicating the number of octets that the sender can transmit in addition to the existing flow control window. The legal range for the increment to the flow control window is 1 to 231-1 (2,147,483,647) octets.

+
+
+

The WINDOW_UPDATE frame does not define any flags.

+
+
+

The WINDOW_UPDATE frame can be specific to a stream or to the entire connection. In the former case, the frame's stream identifier indicates the affected stream; in the latter, the value "0" indicates that the entire connection is the subject of the frame.

+
+
+

A receiver MUST treat the receipt of a WINDOW_UPDATE frame with an flow control window increment of 0 as a stream error (Section 5.4.2) of type PROTOCOL_ERROR; errors on the connection flow control window MUST be treated as a connection error (Section 5.4.1).

+
+
+

WINDOW_UPDATE can be sent by a peer that has sent a frame bearing the END_STREAM flag. This means that a receiver could receive a WINDOW_UPDATE frame on a "half closed (remote)" or "closed" stream. A receiver MUST NOT treat this as an error, see Section 5.1.

+
+
+

A receiver that receives a flow controlled frame MUST always account for its contribution against the connection flow control window, unless the receiver treats this as a connection error (Section 5.4.1). This is necessary even if the frame is in error. Since the sender counts the frame toward the flow control window, if the receiver does not, the flow control window at sender and receiver can become different.

+
+
+

A WINDOW_UPDATE frame with a length other than 4 octets MUST be treated as a connection error (Section 5.4.1) of type FRAME_SIZE_ERROR.

+
+
+

6.9.1 The Flow Control Window +

+
+

Flow control in HTTP/2 is implemented using a window kept by each sender on every stream. The flow control window is a simple integer value that indicates how many octets of data the sender is permitted to transmit; as such, its size is a measure of the buffering capacity of the receiver.

+
+
+

Two flow control windows are applicable: the stream flow control window and the connection flow control window. The sender MUST NOT send a flow controlled frame with a length that exceeds the space available in either of the flow control windows advertised by the receiver. Frames with zero length with the END_STREAM flag set (that is, an empty DATA frame) MAY be sent if there is no available space in either flow control window.

+
+
+

For flow control calculations, the 9 octet frame header is not counted.

+
+
+

After sending a flow controlled frame, the sender reduces the space available in both windows by the length of the transmitted frame.

+
+
+

The receiver of a frame sends a WINDOW_UPDATE frame as it consumes data and frees up space in flow control windows. Separate WINDOW_UPDATE frames are sent for the stream and connection level flow control windows.

+
+
+

A sender that receives a WINDOW_UPDATE frame updates the corresponding window by the amount specified in the frame.

+
+
+

A sender MUST NOT allow a flow control window to exceed 231-1 octets. If a sender receives a WINDOW_UPDATE that causes a flow control window to exceed this maximum it MUST terminate either the stream or the connection, as appropriate. For streams, the sender sends a RST_STREAM with the error code of FLOW_CONTROL_ERROR code; for the connection, a GOAWAY frame with a FLOW_CONTROL_ERROR code.

+
+
+

Flow controlled frames from the sender and WINDOW_UPDATE frames from the receiver are completely asynchronous with respect to each other. This property allows a receiver to aggressively update the window size kept by the sender to prevent streams from stalling.

+
+
+
+

6.9.2 Initial Flow Control Window Size

+
+

When an HTTP/2 connection is first established, new streams are created with an initial flow control window size of 65,535 octets. The connection flow control window is 65,535 octets. Both endpoints can adjust the initial window size for new streams by including a value for SETTINGS_INITIAL_WINDOW_SIZE in the SETTINGS frame that forms part of the connection preface. The connection flow control window can only be changed using WINDOW_UPDATE frames.

+
+
+

Prior to receiving a SETTINGS frame that sets a value for SETTINGS_INITIAL_WINDOW_SIZE, an endpoint can only use the default initial window size when sending flow controlled frames. Similarly, the connection flow control window is set to the default initial window size until a WINDOW_UPDATE frame is received.

+
+
+

A SETTINGS frame can alter the initial flow control window size for all streams in the "open" or "half closed (remote)" state. When the value of SETTINGS_INITIAL_WINDOW_SIZE changes, a receiver MUST adjust the size of all stream flow control windows that it maintains by the difference between the new value and the old value.

+
+
+

A change to SETTINGS_INITIAL_WINDOW_SIZE can cause the available space in a flow control window to become negative. A sender MUST track the negative flow control window, and MUST NOT send new flow controlled frames until it receives WINDOW_UPDATE frames that cause the flow control window to become positive.

+
+
+

For example, if the client sends 60KB immediately on connection establishment, and the server sets the initial window size to be 16KB, the client will recalculate the available flow control window to be -44KB on receipt of the SETTINGS frame. The client retains a negative flow control window until WINDOW_UPDATE frames restore the window to being positive, after which the client can resume sending.

+
+
+

A SETTINGS frame cannot alter the connection flow control window.

+
+
+

An endpoint MUST treat a change to SETTINGS_INITIAL_WINDOW_SIZE that causes any flow control window to exceed the maximum size as a connection error (Section 5.4.1) of type FLOW_CONTROL_ERROR.

+
+
+
+

6.9.3 Reducing the Stream Window Size +

+
+

A receiver that wishes to use a smaller flow control window than the current size can send a new SETTINGS frame. However, the receiver MUST be prepared to receive data that exceeds this window size, since the sender might send data that exceeds the lower limit prior to processing the SETTINGS frame.

+
+
+

After sending a SETTINGS frame that reduces the initial flow control window size, a receiver MAY continue to process streams that exceed flow control limits. Allowing streams to continue does not allow the receiver to immediately reduce the space it reserves for flow control windows. Progress on these streams can also stall, since WINDOW_UPDATE frames are needed to allow the sender to resume sending. The receiver MAY instead send a RST_STREAM with FLOW_CONTROL_ERROR error code for the affected streams.

+
+
+
+
+

6.10 CONTINUATION

+
+

The CONTINUATION frame (type=0x9) is used to continue a sequence of header block fragments (Section 4.3). Any number of CONTINUATION frames can be sent, as long as the preceding frame is on the same stream and is a HEADERS, PUSH_PROMISE or CONTINUATION frame without the END_HEADERS flag set.

+
+
+
 +---------------------------------------------------------------+
+ |                   Header Block Fragment (*)                 ...
+ +---------------------------------------------------------------+
+
+

Figure 15: CONTINUATION Frame Payload

+
+

The CONTINUATION frame payload contains a header block fragment (Section 4.3).

+
+
+

The CONTINUATION frame defines the following flag:

+
+
END_HEADERS (0x4):
+
+

Bit 2 being set indicates that this frame ends a header block (Section 4.3).

+

If the END_HEADERS bit is not set, this frame MUST be followed by another CONTINUATION frame. A receiver MUST treat the receipt of any other type of frame or a frame on a different stream as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
+
+
+
+

The CONTINUATION frame changes the connection state as defined in Section 4.3.

+
+
+

CONTINUATION frames MUST be associated with a stream. If a CONTINUATION frame is received whose stream identifier field is 0x0, the recipient MUST respond with a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
+
+

A CONTINUATION frame MUST be preceded by a HEADERS, PUSH_PROMISE or CONTINUATION frame without the END_HEADERS flag set. A recipient that observes violation of this rule MUST respond with a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
+
+
+
+

7. Error Codes

+
+

Error codes are 32-bit fields that are used in RST_STREAM and GOAWAY frames to convey the reasons for the stream or connection error.

+
+
+

Error codes share a common code space. Some error codes apply only to either streams or the entire connection and have no defined semantics in the other context.

+
+
+

The following error codes are defined:

+
+
NO_ERROR (0x0):
+
The associated condition is not as a result of an error. For example, a GOAWAY might include this code to indicate graceful shutdown of a connection.
+
PROTOCOL_ERROR (0x1):
+
The endpoint detected an unspecific protocol error. This error is for use when a more specific error code is not available.
+
INTERNAL_ERROR (0x2):
+
The endpoint encountered an unexpected internal error.
+
FLOW_CONTROL_ERROR (0x3):
+
The endpoint detected that its peer violated the flow control protocol.
+
SETTINGS_TIMEOUT (0x4):
+
The endpoint sent a SETTINGS frame, but did not receive a response in a timely manner. See Settings Synchronization (Section 6.5.3).
+
STREAM_CLOSED (0x5):
+
The endpoint received a frame after a stream was half closed.
+
FRAME_SIZE_ERROR (0x6):
+
The endpoint received a frame with an invalid size.
+
REFUSED_STREAM (0x7):
+
The endpoint refuses the stream prior to performing any application processing, see Section 8.1.4 for details.
+
CANCEL (0x8):
+
Used by the endpoint to indicate that the stream is no longer needed.
+
COMPRESSION_ERROR (0x9):
+
The endpoint is unable to maintain the header compression context for the connection.
+
CONNECT_ERROR (0xa):
+
The connection established in response to a CONNECT request (Section 8.3) was reset or abnormally closed.
+
ENHANCE_YOUR_CALM (0xb):
+
The endpoint detected that its peer is exhibiting a behavior that might be generating excessive load.
+
INADEQUATE_SECURITY (0xc):
+
The underlying transport has properties that do not meet minimum security requirements (see Section 9.2).
+
HTTP_1_1_REQUIRED (0xd):
+
The endpoint requires that HTTP/1.1 be used instead of HTTP/2.
+
+
+
+

Unknown or unsupported error codes MUST NOT trigger any special behavior. These MAY be treated by an implementation as being equivalent to INTERNAL_ERROR.

+
+
+
+

8. HTTP Message Exchanges

+
+

HTTP/2 is intended to be as compatible as possible with current uses of HTTP. This means that, from the application perspective, the features of the protocol are largely unchanged. To achieve this, all request and response semantics are preserved, although the syntax of conveying those semantics has changed.

+
+
+

Thus, the specification and requirements of HTTP/1.1 Semantics and Content + + [RFC7231] + , Conditional Requests + + [RFC7232] + , Range Requests + + [RFC7233] + , Caching + + [RFC7234] + and Authentication + + [RFC7235] + are applicable to HTTP/2. Selected portions of HTTP/1.1 Message Syntax and Routing + + [RFC7230] + , such as the HTTP and HTTPS URI schemes, are also applicable in HTTP/2, but the expression of those semantics for this protocol are defined in the sections below.

+
+
+

8.1 HTTP Request/Response Exchange

+
+

A client sends an HTTP request on a new stream, using a previously unused stream identifier (Section 5.1.1). A server sends an HTTP response on the same stream as the request.

+
+
+

An HTTP message (request or response) consists of:

+
    +
  1. for a response only, zero or more HEADERS frames (each followed by zero or more CONTINUATION frames) containing the message headers of informational (1xx) HTTP responses (see + + [RFC7230] + , Section 3.2 and + + [RFC7231] + , Section 6.2), and
  2. +
  3. one HEADERS frame (followed by zero or more CONTINUATION frames) containing the message headers (see + + [RFC7230] + , Section 3.2), and
  4. +
  5. zero or more DATA frames containing the payload body (see + + [RFC7230] + , Section 3.3), and
  6. +
  7. optionally, one HEADERS frame, followed by zero or more CONTINUATION frames containing the trailer-part, if present (see + + [RFC7230] + , Section 4.1.2).
  8. +
+

The last frame in the sequence bears an END_STREAM flag, noting that a HEADERS frame bearing the END_STREAM flag can be followed by CONTINUATION frames that carry any remaining portions of the header block.

+
+
+

Other frames (from any stream) MUST NOT occur between either HEADERS frame and any CONTINUATION frames that might follow.

+
+
+

HTTP/2 uses DATA frames to carry message payloads. The + chunked transfer encoding defined in Section 4.1 of + + [RFC7230] + MUST NOT be used in HTTP/2.

+
+
+

Trailing header fields are carried in a header block that also terminates the stream. Such a header block is a sequence starting with a HEADERS frame, followed by zero or more CONTINUATION frames, where the HEADERS frame bears an END_STREAM flag. Header blocks after the first that do not terminate the stream are not part of an HTTP request or response.

+
+
+

A HEADERS frame (and associated CONTINUATION frames) can only appear at the start or end of a stream. An endpoint that receives a HEADERS frame without the END_STREAM flag set after receiving a final (non-informational) status code MUST treat the corresponding request or response as malformed (Section 8.1.2.6).

+
+
+

An HTTP request/response exchange fully consumes a single stream. A request starts with the HEADERS frame that puts the stream into an "open" state. The request ends with a frame bearing END_STREAM, which causes the stream to become "half closed (local)" for the client and "half closed (remote)" for the server. A response starts with a HEADERS frame and ends with a frame bearing END_STREAM, which places the stream in the "closed" state.

+
+
+

An HTTP response is complete after the server sends - or the client receives - a frame with the END_STREAM flag set (including any CONTINUATION frames needed to complete a header block). A server can send a complete response prior to the client sending an entire request if the response does not depend on any portion of the request that has not been sent and received. When this is true, a server MAY request that the client abort transmission of a request without error by sending a RST_STREAM with an error code of NO_ERROR after sending a complete response (i.e., a frame with the END_STREAM flag). Clients MUST NOT discard responses as a result of receiving such a RST_STREAM, though clients can always discard responses at their discretion for other reasons.

+
+
+

8.1.1 Upgrading From HTTP/2

+
+

HTTP/2 removes support for the 101 (Switching Protocols) informational status code ( + + [RFC7231] + , Section 6.2.2).

+
+
+

The semantics of 101 (Switching Protocols) aren't applicable to a multiplexed protocol. Alternative protocols are able to use the same mechanisms that HTTP/2 uses to negotiate their use (see Section 3).

+
+
+
+

8.1.2 HTTP Header Fields

+
+

HTTP header fields carry information as a series of key-value pairs. For a listing of registered HTTP headers, see the Message Header Field Registry maintained at <https://www.iana.org/assignments/message-headers>.

+
+
+

Just as in HTTP/1.x, header field names are strings of ASCII characters that are compared in a case-insensitive fashion. However, header field names MUST be converted to lowercase prior to their encoding in HTTP/2. A request or response containing uppercase header field names MUST be treated as malformed (Section 8.1.2.6).

+
+
+

8.1.2.1 Pseudo-Header Fields

+
+

While HTTP/1.x used the message start-line (see + + [RFC7230] + , Section 3.1) to convey the target URI and method of the request, and the status code for the response, HTTP/2 uses special pseudo-header fields beginning with ':' character (ASCII 0x3a) for this purpose.

+
+
+

Pseudo-header fields are not HTTP header fields. Endpoints MUST NOT generate pseudo-header fields other than those defined in this document.

+
+
+

Pseudo-header fields are only valid in the context in which they are defined. Pseudo-header fields defined for requests MUST NOT appear in responses; pseudo-header fields defined for responses MUST NOT appear in requests. Pseudo-header fields MUST NOT appear in trailers. Endpoints MUST treat a request or response that contains undefined or invalid pseudo-header fields as malformed (Section 8.1.2.6).

+
+
+

All pseudo-header fields MUST appear in the header block before regular header fields. Any request or response that contains a pseudo-header field that appears in a header block after a regular header field MUST be treated as malformed (Section 8.1.2.6).

+
+
+
+

8.1.2.2 Connection-Specific Header Fields +

+
+

HTTP/2 does not use the + Connection header field to indicate connection-specific header fields; in this protocol, connection-specific metadata is conveyed by other means. An endpoint MUST NOT generate an HTTP/2 message containing connection-specific header fields; any message containing connection-specific header fields MUST be treated as malformed (Section 8.1.2.6).

+
+
+

The only exception to this is the TE header field, which MAY be present in an HTTP/2 request; when it is, it MUST NOT contain any value other than "trailers".

+
+
+

This means that an intermediary transforming an HTTP/1.x message to HTTP/2 will need to remove any header fields nominated by the Connection header field, along with the Connection header field itself. Such intermediaries SHOULD also remove other connection-specific header fields, such as Keep-Alive, Proxy-Connection, Transfer-Encoding and Upgrade, even if they are not nominated by Connection.

+
+
+
+
Note:
+
HTTP/2 purposefully does not support upgrade to another protocol. The handshake methods described in Section 3 are believed sufficient to negotiate the use of alternative protocols.
+
+
+
+
+

8.1.2.3 Request Pseudo-Header Fields

+
+

The following pseudo-header fields are defined for HTTP/2 requests:

+
    +
  • +

    The + :method pseudo-header field includes the HTTP method ( + + [RFC7231] + , Section 4).

    +
  • +
  • +

    The + :scheme pseudo-header field includes the scheme portion of the target URI ( + + [RFC3986] + , Section 3.1).

    +

    + :scheme is not restricted to + http and + https schemed URIs. A proxy or gateway can translate requests for non-HTTP schemes, enabling the use of HTTP to interact with non-HTTP services.

    +
  • +
  • +

    The + :authority pseudo-header field includes the authority portion of the target URI ( + + [RFC3986] + , Section 3.2). The authority MUST NOT include the deprecated + userinfo subcomponent for + http or + https schemed URIs.

    +

    To ensure that the HTTP/1.1 request line can be reproduced accurately, this pseudo-header field MUST be omitted when translating from an HTTP/1.1 request that has a request target in origin or asterisk form (see + + [RFC7230] + , Section 5.3). Clients that generate HTTP/2 requests directly SHOULD use the + :authority pseudo-header field instead of the + Host header field. An intermediary that converts an HTTP/2 request to HTTP/1.1 MUST create a + Host header field if one is not present in a request by copying the value of the + :authority pseudo-header field.

    +
  • +
  • +

    The + :path pseudo-header field includes the path and query parts of the target URI (the + path-absolute production from + + [RFC3986] + and optionally a '?' character followed by the + query production, see + + [RFC3986] + , Section 3.3 and + + [RFC3986] + , Section 3.4). A request in asterisk form includes the value '*' for the + :path pseudo-header field.

    +

    This pseudo-header field MUST NOT be empty for + http or + https URIs; + http or + https URIs that do not contain a path component MUST include a value of '/'. The exception to this rule is an OPTIONS request for an + http or + https URI that does not include a path component; these MUST include a + :path pseudo-header field with a value of '*' (see + + [RFC7230] + , Section 5.3.4).

    +
  • +
+
+
+

All HTTP/2 requests MUST include exactly one valid value for the + :method, + :scheme, and + :path pseudo-header fields, unless it is a CONNECT request (Section 8.3). An HTTP request that omits mandatory pseudo-header fields is malformed (Section 8.1.2.6).

+
+
+

HTTP/2 does not define a way to carry the version identifier that is included in the HTTP/1.1 request line.

+
+
+
+

8.1.2.4 Response Pseudo-Header Fields

+
+

For HTTP/2 responses, a single + :status pseudo-header field is defined that carries the HTTP status code field (see + + [RFC7231] + , Section 6). This pseudo-header field MUST be included in all responses, otherwise the response is malformed (Section 8.1.2.6).

+
+
+

HTTP/2 does not define a way to carry the version or reason phrase that is included in an HTTP/1.1 status line.

+
+
+
+

8.1.2.5 Compressing the Cookie Header Field

+
+

The Cookie header field + [COOKIE] uses a semi-colon (";") to delimit cookie-pairs (or "crumbs"). This header field doesn't follow the list construction rules in HTTP (see + + [RFC7230] + , Section 3.2.2), which prevents cookie-pairs from being separated into different name-value pairs. This can significantly reduce compression efficiency as individual cookie-pairs are updated.

+
+
+

To allow for better compression efficiency, the Cookie header field MAY be split into separate header fields, each with one or more cookie-pairs. If there are multiple Cookie header fields after decompression, these MUST be concatenated into a single octet string using the two octet delimiter of 0x3B, 0x20 (the ASCII string "; ") before being passed into a non-HTTP/2 context, such as an HTTP/1.1 connection, or a generic HTTP server application.

+
+
+

Therefore, the following two lists of Cookie header fields are semantically equivalent.

  cookie: a=b; c=d; e=f
+
+  cookie: a=b
+  cookie: c=d
+  cookie: e=f
+
+
+
+

8.1.2.6 Malformed Requests and Responses

+
+

A malformed request or response is one that is an otherwise valid sequence of HTTP/2 frames, but is otherwise invalid due to the presence of extraneous frames, prohibited header fields, the absence of mandatory header fields, or the inclusion of uppercase header field names.

+
+
+

A request or response that includes an payload body can include a + content-length header field. A request or response is also malformed if the value of a + content-length header field does not equal the sum of the DATA frame payload lengths that form the body. A response that is defined to have no payload, as described in + + [RFC7230] + , Section 3.3.2, can have a non-zero + content-length header field, even though no content is included in DATA frames.

+
+
+

Intermediaries that process HTTP requests or responses (i.e., any intermediary not acting as a tunnel) MUST NOT forward a malformed request or response. Malformed requests or responses that are detected MUST be treated as a stream error (Section 5.4.2) of type PROTOCOL_ERROR.

+
+
+

For malformed requests, a server MAY send an HTTP response prior to closing or resetting the stream. Clients MUST NOT accept a malformed response. Note that these requirements are intended to protect against several types of common attacks against HTTP; they are deliberately strict, because being permissive can expose implementations to these vulnerabilities.

+
+
+
+
+

8.1.3 Examples +

+
+

This section shows HTTP/1.1 requests and responses, with illustrations of equivalent HTTP/2 requests and responses.

+
+
+

An HTTP GET request includes request header fields and no payload body and is therefore transmitted as a single HEADERS frame, followed by zero or more CONTINUATION frames containing the serialized block of request header fields. The HEADERS frame in the following has both the END_HEADERS and END_STREAM flags set; no CONTINUATION frames are sent:

+
+
  GET /resource HTTP/1.1           HEADERS
+  Host: example.org          ==>     + END_STREAM
+  Accept: image/jpeg                 + END_HEADERS
+                                       :method = GET
+                                       :scheme = https
+                                       :path = /resource
+                                       host = example.org
+                                       accept = image/jpeg
+
+
+

Similarly, a response that includes only response header fields is transmitted as a HEADERS frame (again, followed by zero or more CONTINUATION frames) containing the serialized block of response header fields.

+
+
  HTTP/1.1 304 Not Modified        HEADERS
+  ETag: "xyzzy"              ==>     + END_STREAM
+  Expires: Thu, 23 Jan ...           + END_HEADERS
+                                       :status = 304
+                                       etag = "xyzzy"
+                                       expires = Thu, 23 Jan ...
+
+
+

An HTTP POST request that includes request header fields and payload data is transmitted as one HEADERS frame, followed by zero or more CONTINUATION frames containing the request header fields, followed by one or more DATA frames, with the last CONTINUATION (or HEADERS) frame having the END_HEADERS flag set and the final DATA frame having the END_STREAM flag set:

+
+
  POST /resource HTTP/1.1          HEADERS
+  Host: example.org          ==>     - END_STREAM
+  Content-Type: image/jpeg           - END_HEADERS
+  Content-Length: 123                  :method = POST
+                                       :path = /resource
+  {binary data}                        :scheme = https
+
+                                   CONTINUATION
+                                     + END_HEADERS
+                                       content-type = image/jpeg
+                                       host = example.org
+                                       content-length = 123
+
+                                   DATA
+                                     + END_STREAM
+                                   {binary data}
+
+

Note that data contributing to any given header field could be spread between header block fragments. The allocation of header fields to frames in this example is illustrative only.

+
+
+

A response that includes header fields and payload data is transmitted as a HEADERS frame, followed by zero or more CONTINUATION frames, followed by one or more DATA frames, with the last DATA frame in the sequence having the END_STREAM flag set:

+
+
  HTTP/1.1 200 OK                  HEADERS
+  Content-Type: image/jpeg   ==>     - END_STREAM
+  Content-Length: 123                + END_HEADERS
+                                       :status = 200
+  {binary data}                        content-type = image/jpeg
+                                       content-length = 123
+
+                                   DATA
+                                     + END_STREAM
+                                   {binary data}
+
+
+

An informational response using a 1xx status code other than 101 is transmitted as a HEADERS frame, followed by zero or more CONTINUATION frames.

+
+
+

Trailing header fields are sent as a header block after both the request or response header block and all the DATA frames have been sent. The HEADERS frame starting the trailers header block has the END_STREAM flag set.

+
+
+

The following example includes both a 100 (Continue) status code, which is sent in response to a request containing a "100-continue" token in the Expect header field, and trailing header fields:

  HTTP/1.1 100 Continue            HEADERS
+  Extension-Field: bar       ==>     - END_STREAM
+                                     + END_HEADERS
+                                       :status = 100
+                                       extension-field = bar
+
+  HTTP/1.1 200 OK                  HEADERS
+  Content-Type: image/jpeg   ==>     - END_STREAM
+  Transfer-Encoding: chunked         + END_HEADERS
+  Trailer: Foo                         :status = 200
+                                       content-length = 123
+  123                                  content-type = image/jpeg
+  {binary data}                        trailer = Foo
+  0
+  Foo: bar                         DATA
+                                     - END_STREAM
+                                   {binary data}
+
+                                   HEADERS
+                                     + END_STREAM
+                                     + END_HEADERS
+                                       foo = bar
+
+
+
+

8.1.4 Request Reliability Mechanisms in HTTP/2

+
+

In HTTP/1.1, an HTTP client is unable to retry a non-idempotent request when an error occurs, because there is no means to determine the nature of the error. It is possible that some server processing occurred prior to the error, which could result in undesirable effects if the request were reattempted.

+
+
+

HTTP/2 provides two mechanisms for providing a guarantee to a client that a request has not been processed:

+
    +
  • The GOAWAY frame indicates the highest stream number that might have been processed. Requests on streams with higher numbers are therefore guaranteed to be safe to retry.
  • +
  • The REFUSED_STREAM error code can be included in a RST_STREAM frame to indicate that the stream is being closed prior to any processing having occurred. Any request that was sent on the reset stream can be safely retried.
  • +
+
+
+

Requests that have not been processed have not failed; clients MAY automatically retry them, even those with non-idempotent methods.

+
+
+

A server MUST NOT indicate that a stream has not been processed unless it can guarantee that fact. If frames that are on a stream are passed to the application layer for any stream, then REFUSED_STREAM MUST NOT be used for that stream, and a GOAWAY frame MUST include a stream identifier that is greater than or equal to the given stream identifier.

+
+
+

In addition to these mechanisms, the PING frame provides a way for a client to easily test a connection. Connections that remain idle can become broken as some middleboxes (for instance, network address translators, or load balancers) silently discard connection bindings. The PING frame allows a client to safely test whether a connection is still active without sending a request.

+
+
+
+
+

8.2 Server Push

+
+

HTTP/2 allows a server to pre-emptively send (or "push") responses (along with corresponding "promised" requests) to a client in association with a previous client-initiated request. This can be useful when the server knows the client will need to have those responses available in order to fully process the response to the original request.

+
+
+

A client can request that server push be disabled, though this is negotiated for each hop independently. The SETTINGS_ENABLE_PUSH setting can be set to 0 to indicate that server push is disabled.

+
+
+

Promised requests MUST be cacheable (see + + [RFC7231] + , Section 4.2.3), MUST be safe (see + + [RFC7231] + , Section 4.2.1) and MUST NOT include a request body. Clients that receive a promised request that is not cacheable, is not known to be safe or that indicates the presence of a request body MUST reset the promised stream with a stream error (Section 5.4.2) of type PROTOCOL_ERROR. Note this could result in the promised stream being reset if the client does not recognize a newly defined method as being safe.

+
+
+

Pushed responses that are cacheable (see + + [RFC7234] + , Section 3) can be stored by the client, if it implements an HTTP cache. Pushed responses are considered successfully validated on the origin server (e.g., if the "no-cache" cache response directive + + [RFC7234] + , Section 5.2.2 is present) while the stream identified by the promised stream ID is still open.

+
+
+

Pushed responses that are not cacheable MUST NOT be stored by any HTTP cache. They MAY be made available to the application separately.

+
+
+

The server MUST include a value in the + :authority header field for which the server is authoritative (see Section 10.1). A client MUST treat a PUSH_PROMISE for which the server is not authoritative as a stream error (Section 5.4.2) of type PROTOCOL_ERROR.

+
+
+

An intermediary can receive pushes from the server and choose not to forward them on to the client. In other words, how to make use of the pushed information is up to that intermediary. Equally, the intermediary might choose to make additional pushes to the client, without any action taken by the server.

+
+
+

A client cannot push. Thus, servers MUST treat the receipt of a PUSH_PROMISE frame as a connection error (Section 5.4.1) of type PROTOCOL_ERROR. Clients MUST reject any attempt to change the SETTINGS_ENABLE_PUSH setting to a value other than 0 by treating the message as a connection error (Section 5.4.1) of type PROTOCOL_ERROR.

+
+
+

8.2.1 Push Requests

+
+

Server push is semantically equivalent to a server responding to a request; however, in this case that request is also sent by the server, as a PUSH_PROMISE frame.

+
+
+

The PUSH_PROMISE frame includes a header block that contains a complete set of request header fields that the server attributes to the request. It is not possible to push a response to a request that includes a request body.

+
+
+

Pushed responses are always associated with an explicit request from the client. The PUSH_PROMISE frames sent by the server are sent on that explicit request's stream. The PUSH_PROMISE frame also includes a promised stream identifier, chosen from the stream identifiers available to the server (see Section 5.1.1).

+
+
+

The header fields in PUSH_PROMISE and any subsequent CONTINUATION frames MUST be a valid and complete set of request header fields (Section 8.1.2.3). The server MUST include a method in the + :method header field that is safe and cacheable. If a client receives a PUSH_PROMISE that does not include a complete and valid set of header fields, or the + :method header field identifies a method that is not safe, it MUST respond with a stream error (Section 5.4.2) of type PROTOCOL_ERROR.

+
+
+

The server SHOULD send PUSH_PROMISE (Section 6.6) frames prior to sending any frames that reference the promised responses. This avoids a race where clients issue requests prior to receiving any PUSH_PROMISE frames.

+
+
+

For example, if the server receives a request for a document containing embedded links to multiple image files, and the server chooses to push those additional images to the client, sending push promises before the DATA frames that contain the image links ensures that the client is able to see the promises before discovering embedded links. Similarly, if the server pushes responses referenced by the header block (for instance, in Link header fields), sending the push promises before sending the header block ensures that clients do not request them.

+
+
+

PUSH_PROMISE frames MUST NOT be sent by the client.

+
+
+

PUSH_PROMISE frames can be sent by the server in response to any client-initiated stream, but the stream MUST be in either the "open" or "half closed (remote)" state with respect to the server. PUSH_PROMISE frames are interspersed with the frames that comprise a response, though they cannot be interspersed with HEADERS and CONTINUATION frames that comprise a single header block.

+
+
+

Sending a PUSH_PROMISE frame creates a new stream and puts the stream into the “reserved (local)” state for the server and the “reserved (remote)” state for the client.

+
+
+
+

8.2.2 Push Responses

+
+

After sending the PUSH_PROMISE frame, the server can begin delivering the pushed response as a response (Section 8.1.2.4) on a server-initiated stream that uses the promised stream identifier. The server uses this stream to transmit an HTTP response, using the same sequence of frames as defined in Section 8.1. This stream becomes "half closed" to the client (Section 5.1) after the initial HEADERS frame is sent.

+
+
+

Once a client receives a PUSH_PROMISE frame and chooses to accept the pushed response, the client SHOULD NOT issue any requests for the promised response until after the promised stream has closed.

+
+
+

If the client determines, for any reason, that it does not wish to receive the pushed response from the server, or if the server takes too long to begin sending the promised response, the client can send an RST_STREAM frame, using either the CANCEL or REFUSED_STREAM codes, and referencing the pushed stream's identifier.

+
+
+

A client can use the SETTINGS_MAX_CONCURRENT_STREAMS setting to limit the number of responses that can be concurrently pushed by a server. Advertising a SETTINGS_MAX_CONCURRENT_STREAMS value of zero disables server push by preventing the server from creating the necessary streams. This does not prohibit a server from sending PUSH_PROMISE frames; clients need to reset any promised streams that are not wanted.

+
+
+

Clients receiving a pushed response MUST validate that either the server is authoritative (see Section 10.1), or the proxy that provided the pushed response is configured for the corresponding request. For example, a server that offers a certificate for only the + example.com DNS-ID or Common Name is not permitted to push a response for + https://www.example.org/doc.

+
+
+

The response for a PUSH_PROMISE stream begins with a HEADERS frame, which immediately puts the stream into the “half closed (remote)” state for the server and “half closed (local)” state for the client, and ends with a frame bearing END_STREAM, which places the stream in the "closed" state.

+
+
Note:
+
The client never sends a frame with the END_STREAM flag for a server push.
+
+
+
+
+
+

8.3 The CONNECT Method

+
+

In HTTP/1.x, the pseudo-method CONNECT ( + + [RFC7231] + , Section 4.3.6) is used to convert an HTTP connection into a tunnel to a remote host. CONNECT is primarily used with HTTP proxies to establish a TLS session with an origin server for the purposes of interacting with + https resources.

+
+
+

In HTTP/2, the CONNECT method is used to establish a tunnel over a single HTTP/2 stream to a remote host, for similar purposes. The HTTP header field mapping works as defined in Request Header Fields (Section 8.1.2.3), with a few differences. Specifically:

+
    +
  • The + :method header field is set to + CONNECT.
  • +
  • The + :scheme and + :path header fields MUST be omitted.
  • +
  • The + :authority header field contains the host and port to connect to (equivalent to the authority-form of the request-target of CONNECT requests, see + + [RFC7230] + , Section 5.3).
  • +
+
+
+

A CONNECT request that does not conform to these restrictions is malformed (Section 8.1.2.6).

+
+
+

A proxy that supports CONNECT establishes a TCP connection + [TCP] to the server identified in the + :authority header field. Once this connection is successfully established, the proxy sends a HEADERS frame containing a 2xx series status code to the client, as defined in + + [RFC7231] + , Section 4.3.6.

+
+
+

After the initial HEADERS frame sent by each peer, all subsequent DATA frames correspond to data sent on the TCP connection. The payload of any DATA frames sent by the client is transmitted by the proxy to the TCP server; data received from the TCP server is assembled into DATA frames by the proxy. Frame types other than DATA or stream management frames (RST_STREAM, WINDOW_UPDATE, and PRIORITY) MUST NOT be sent on a connected stream, and MUST be treated as a stream error (Section 5.4.2) if received.

+
+
+

The TCP connection can be closed by either peer. The END_STREAM flag on a DATA frame is treated as being equivalent to the TCP FIN bit. A client is expected to send a DATA frame with the END_STREAM flag set after receiving a frame bearing the END_STREAM flag. A proxy that receives a DATA frame with the END_STREAM flag set sends the attached data with the FIN bit set on the last TCP segment. A proxy that receives a TCP segment with the FIN bit set sends a DATA frame with the END_STREAM flag set. Note that the final TCP segment or DATA frame could be empty.

+
+
+

A TCP connection error is signaled with RST_STREAM. A proxy treats any error in the TCP connection, which includes receiving a TCP segment with the RST bit set, as a stream error (Section 5.4.2) of type CONNECT_ERROR. Correspondingly, a proxy MUST send a TCP segment with the RST bit set if it detects an error with the stream or the HTTP/2 connection.

+
+
+
+
+

9. Additional HTTP Requirements/Considerations

+
+

This section outlines attributes of the HTTP protocol that improve interoperability, reduce exposure to known security vulnerabilities, or reduce the potential for implementation variation.

+
+
+

9.1 Connection Management +

+
+

HTTP/2 connections are persistent. For best performance, it is expected clients will not close connections until it is determined that no further communication with a server is necessary (for example, when a user navigates away from a particular web page), or until the server closes the connection.

+
+
+

Clients SHOULD NOT open more than one HTTP/2 connection to a given host and port pair, where host is derived from a URI, a selected alternative service + [ALT-SVC], or a configured proxy.

+
+
+

A client can create additional connections as replacements, either to replace connections that are near to exhausting the available stream identifier space (Section 5.1.1), to refresh the keying material for a TLS connection, or to replace connections that have encountered errors (Section 5.4.1).

+
+
+

A client MAY open multiple connections to the same IP address and TCP port using different Server Name Indication + [TLS-EXT] values or to provide different TLS client certificates, but SHOULD avoid creating multiple connections with the same configuration.

+
+
+

Servers are encouraged to maintain open connections for as long as possible, but are permitted to terminate idle connections if necessary. When either endpoint chooses to close the transport-layer TCP connection, the terminating endpoint SHOULD first send a GOAWAY (Section 6.8) frame so that both endpoints can reliably determine whether previously sent frames have been processed and gracefully complete or terminate any necessary remaining tasks.

+
+
+

9.1.1 Connection Reuse

+
+

Connections that are made to an origin server, either directly or through a tunnel created using the CONNECT method (Section 8.3) MAY be reused for requests with multiple different URI authority components. A connection can be reused as long as the origin server is authoritative (Section 10.1). For TCP connections without TLS, this depends on the host having resolved to the same IP address.

+
+
+

For + https resources, connection reuse additionally depends on having a certificate that is valid for the host in the URI. The certificate presented by the server MUST satisfy any checks that the client would perform when forming a new TLS connection for the host in the URI.

+
+
+

An origin server might offer a certificate with multiple + subjectAltName attributes, or names with wildcards, one of which is valid for the authority in the URI. For example, a certificate with a + subjectAltName of + *.example.com might permit the use of the same connection for requests to URIs starting with + https://a.example.com/ and + https://b.example.com/.

+
+
+

In some deployments, reusing a connection for multiple origins can result in requests being directed to the wrong origin server. For example, TLS termination might be performed by a middlebox that uses the TLS Server Name Indication (SNI) + [TLS-EXT] extension to select an origin server. This means that it is possible for clients to send confidential information to servers that might not be the intended target for the request, even though the server is otherwise authoritative.

+
+
+

A server that does not wish clients to reuse connections can indicate that it is not authoritative for a request by sending a 421 (Misdirected Request) status code in response to the request (see Section 9.1.2).

+
+
+

A client that is configured to use a proxy over HTTP/2 directs requests to that proxy through a single connection. That is, all requests sent via a proxy reuse the connection to the proxy.

+
+
+
+

9.1.2 The 421 (Misdirected Request) Status Code

+
+

The 421 (Misdirected Request) status code indicates that the request was directed at a server that is not able to produce a response. This can be sent by a server that is not configured to produce responses for the combination of scheme and authority that are included in the request URI.

+
+
+

Clients receiving a 421 (Misdirected Request) response from a server MAY retry the request - whether the request method is idempotent or not - over a different connection. This is possible if a connection is reused (Section 9.1.1) or if an alternative service is selected ( + + [ALT-SVC] + ).

+
+
+

This status code MUST NOT be generated by proxies.

+
+
+

A 421 response is cacheable by default; i.e., unless otherwise indicated by the method definition or explicit cache controls (see Section 4.2.2 of + + [RFC7234] + ).

+
+
+
+
+

9.2 Use of TLS Features

+
+

Implementations of HTTP/2 MUST use TLS + [TLS12] version 1.2 or higher for HTTP/2 over TLS. The general TLS usage guidance in + + [TLSBCP] + SHOULD be followed, with some additional restrictions that are specific to HTTP/2.

+
+
+

The TLS implementation MUST support the Server Name Indication (SNI) + [TLS-EXT] extension to TLS. HTTP/2 clients MUST indicate the target domain name when negotiating TLS.

+
+
+

Deployments of HTTP/2 that negotiate TLS 1.3 or higher need only support and use the SNI extension; deployments of TLS 1.2 are subject to the requirements in the following sections. Implementations are encouraged to provide defaults that comply, but it is recognized that deployments are ultimately responsible for compliance.

+
+
+

9.2.1 TLS 1.2 Features +

+
+

This section describes restrictions on the TLS 1.2 feature set that can be used with HTTP/2. Due to deployment limitations, it might not be possible to fail TLS negotiation when these restrictions are not met. An endpoint MAY immediately terminate an HTTP/2 connection that does not meet these TLS requirements with a connection error (Section 5.4.1) of type INADEQUATE_SECURITY.

+
+
+

A deployment of HTTP/2 over TLS 1.2 MUST disable compression. TLS compression can lead to the exposure of information that would not otherwise be revealed + + [RFC3749] + . Generic compression is unnecessary since HTTP/2 provides compression features that are more aware of context and therefore likely to be more appropriate for use for performance, security or other reasons.

+
+
+

A deployment of HTTP/2 over TLS 1.2 MUST disable renegotiation. An endpoint MUST treat a TLS renegotiation as a connection error (Section 5.4.1) of type PROTOCOL_ERROR. Note that disabling renegotiation can result in long-lived connections becoming unusable due to limits on the number of messages the underlying cipher suite can encipher.

+
+
+

An endpoint MAY use renegotiation to provide confidentiality protection for client credentials offered in the handshake, but any renegotiation MUST occur prior to sending the connection preface. A server SHOULD request a client certificate if it sees a renegotiation request immediately after establishing a connection.

+
+
+

This effectively prevents the use of renegotiation in response to a request for a specific protected resource. A future specification might provide a way to support this use case. Alternatively, a server might use an error (Section 5.4) of type HTTP_1_1_REQUIRED to request the client use a protocol which supports renegotiation.

+
+
+

Implementations MUST support ephemeral key exchange sizes of at least 2048 bits for cipher suites that use ephemeral finite field Diffie-Hellman (DHE) + + [TLS12] + and 224 bits for cipher suites that use ephemeral elliptic curve Diffie-Hellman (ECDHE) + + [RFC4492] + . Clients MUST accept DHE sizes of up to 4096 bits. Endpoints MAY treat negotiation of key sizes smaller than the lower limits as a connection error (Section 5.4.1) of type INADEQUATE_SECURITY.

+
+
+
+

9.2.2 TLS 1.2 Cipher Suites +

+
+

A deployment of HTTP/2 over TLS 1.2 SHOULD NOT use any of the cipher suites that are listed in the cipher suite black list (Appendix A).

+
+
+

Endpoints MAY choose to generate a connection error (Section 5.4.1) of type INADEQUATE_SECURITY if one of the cipher suites from the black list are negotiated. A deployment that chooses to use a black-listed cipher suite risks triggering a connection error unless the set of potential peers is known to accept that cipher suite.

+
+
+

Implementations MUST NOT generate this error in reaction to the negotiation of a cipher suite that is not on the black list. Consequently, when clients offer a cipher suite that is not on the black list, they have to be prepared to use that cipher suite with HTTP/2.

+
+
+

The black list includes the cipher suite that TLS 1.2 makes mandatory, which means that TLS 1.2 deployments could have non-intersecting sets of permitted cipher suites. To avoid this problem causing TLS handshake failures, deployments of HTTP/2 that use TLS 1.2 MUST support TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256 + + [TLS-ECDHE] + with the P256 elliptic curve + + [FIPS186] + .

+
+
+

Note that clients might advertise support of cipher suites that are on the black list in order to allow for connection to servers that do not support HTTP/2. This allows servers to select HTTP/1.1 with a cipher suite that is on the HTTP/2 black list. However, this can result in HTTP/2 being negotiated with a black-listed cipher suite if the application protocol and cipher suite are independently selected.

+
+
+
+
+
+

10. Security Considerations

+
+

10.1 Server Authority

+
+

HTTP/2 relies on the HTTP/1.1 definition of authority for determining whether a server is authoritative in providing a given response, see + + [RFC7230] + , Section 9.1. This relies on local name resolution for the "http" URI scheme, and the authenticated server identity for the "https" scheme (see + + [RFC2818] + , Section 3).

+
+
+
+

10.2 Cross-Protocol Attacks +

+
+

In a cross-protocol attack, an attacker causes a client to initiate a transaction in one protocol toward a server that understands a different protocol. An attacker might be able to cause the transaction to appear as valid transaction in the second protocol. In combination with the capabilities of the web context, this can be used to interact with poorly protected servers in private networks.

+
+
+

Completing a TLS handshake with an ALPN identifier for HTTP/2 can be considered sufficient protection against cross protocol attacks. ALPN provides a positive indication that a server is willing to proceed with HTTP/2, which prevents attacks on other TLS-based protocols.

+
+
+

The encryption in TLS makes it difficult for attackers to control the data which could be used in a cross-protocol attack on a cleartext protocol.

+
+
+

The cleartext version of HTTP/2 has minimal protection against cross-protocol attacks. The connection preface (Section 3.5) contains a string that is designed to confuse HTTP/1.1 servers, but no special protection is offered for other protocols. A server that is willing to ignore parts of an HTTP/1.1 request containing an Upgrade header field in addition to the client connection preface could be exposed to a cross-protocol attack.

+
+
+
+

10.3 Intermediary Encapsulation Attacks +

+
+

The HTTP/2 header field encoding allows the expression of names that are not valid field names in the Internet Message Syntax used by HTTP/1.1. Requests or responses containing invalid header field names MUST be treated as malformed (Section 8.1.2.6). An intermediary therefore cannot translate an HTTP/2 request or response containing an invalid field name into an HTTP/1.1 message.

+
+
+

Similarly, HTTP/2 allows header field values that are not valid. While most of the values that can be encoded will not alter header field parsing, carriage return (CR, ASCII 0xd), line feed (LF, ASCII 0xa), and the zero character (NUL, ASCII 0x0) might be exploited by an attacker if they are translated verbatim. Any request or response that contains a character not permitted in a header field value MUST be treated as malformed (Section 8.1.2.6). Valid characters are defined by the + field-content ABNF rule in Section 3.2 of + + [RFC7230] + .

+
+
+
+

10.4 Cacheability of Pushed Responses +

+
+

Pushed responses do not have an explicit request from the client; the request is provided by the server in the PUSH_PROMISE frame.

+
+
+

Caching responses that are pushed is possible based on the guidance provided by the origin server in the Cache-Control header field. However, this can cause issues if a single server hosts more than one tenant. For example, a server might offer multiple users each a small portion of its URI space.

+
+
+

Where multiple tenants share space on the same server, that server MUST ensure that tenants are not able to push representations of resources that they do not have authority over. Failure to enforce this would allow a tenant to provide a representation that would be served out of cache, overriding the actual representation that the authoritative tenant provides.

+
+
+

Pushed responses for which an origin server is not authoritative (see Section 10.1) MUST NOT be used or cached.

+
+
+
+

10.5 Denial of Service Considerations

+
+

An HTTP/2 connection can demand a greater commitment of resources to operate than a HTTP/1.1 connection. The use of header compression and flow control depend on a commitment of resources for storing a greater amount of state. Settings for these features ensure that memory commitments for these features are strictly bounded.

+
+
+

The number of PUSH_PROMISE frames is not constrained in the same fashion. A client that accepts server push SHOULD limit the number of streams it allows to be in the "reserved (remote)" state. Excessive number of server push streams can be treated as a stream error (Section 5.4.2) of type ENHANCE_YOUR_CALM.

+
+
+

Processing capacity cannot be guarded as effectively as state capacity.

+
+
+

The SETTINGS frame can be abused to cause a peer to expend additional processing time. This might be done by pointlessly changing SETTINGS parameters, setting multiple undefined parameters, or changing the same setting multiple times in the same frame. WINDOW_UPDATE or PRIORITY frames can be abused to cause an unnecessary waste of resources.

+
+
+

Large numbers of small or empty frames can be abused to cause a peer to expend time processing frame headers. Note however that some uses are entirely legitimate, such as the sending of an empty DATA or CONTINUATION frame at the end of a stream.

+
+
+

Header compression also offers some opportunities to waste processing resources; see Section 7 of + + [COMPRESSION] + for more details on potential abuses.

+
+
+

Limits in SETTINGS parameters cannot be reduced instantaneously, which leaves an endpoint exposed to behavior from a peer that could exceed the new limits. In particular, immediately after establishing a connection, limits set by a server are not known to clients and could be exceeded without being an obvious protocol violation.

+
+
+

All these features - i.e., SETTINGS changes, small frames, header compression - have legitimate uses. These features become a burden only when they are used unnecessarily or to excess.

+
+
+

An endpoint that doesn't monitor this behavior exposes itself to a risk of denial of service attack. Implementations SHOULD track the use of these features and set limits on their use. An endpoint MAY treat activity that is suspicious as a connection error (Section 5.4.1) of type ENHANCE_YOUR_CALM.

+
+
+

10.5.1 Limits on Header Block Size

+
+

A large header block (Section 4.3) can cause an implementation to commit a large amount of state. Header fields that are critical for routing can appear toward the end of a header block, which prevents streaming of header fields to their ultimate destination. This ordering and other reasons, such as ensuring cache correctness, means that an endpoint might need to buffer the entire header block. Since there is no hard limit to the size of a header block, some endpoints could be forced to commit a large amount of available memory for header fields.

+
+
+

An endpoint can use the SETTINGS_MAX_HEADER_LIST_SIZE to advise peers of limits that might apply on the size of header blocks. This setting is only advisory, so endpoints MAY choose to send header blocks that exceed this limit and risk having the request or response being treated as malformed. This setting is specific to a connection, so any request or response could encounter a hop with a lower, unknown limit. An intermediary can attempt to avoid this problem by passing on values presented by different peers, but they are not obligated to do so.

+
+
+

A server that receives a larger header block than it is willing to handle can send an HTTP 431 (Request Header Fields Too Large) status code + + [RFC6585] + . A client can discard responses that it cannot process. The header block MUST be processed to ensure a consistent connection state, unless the connection is closed.

+
+
+
+

10.5.2 CONNECT Issues

+
+

The CONNECT method can be used to create disproportionate load on an proxy, since stream creation is relatively inexpensive when compared to the creation and maintenance of a TCP connection. A proxy might also maintain some resources for a TCP connection beyond the closing of the stream that carries the CONNECT request, since the outgoing TCP connection remains in the TIME_WAIT state. A proxy therefore cannot rely on SETTINGS_MAX_CONCURRENT_STREAMS alone to limit the resources consumed by CONNECT requests.

+
+
+
+
+

10.6 Use of Compression +

+
+

Compression can allow an attacker to recover secret data when it is compressed in the same context as data under attacker control. HTTP/2 enables compression of header fields (Section 4.3); the following concerns also apply to the use of HTTP compressed content-codings ( + + [RFC7231] + , Section 3.1.2.1).

+
+
+

There are demonstrable attacks on compression that exploit the characteristics of the web (e.g., + + [BREACH] + ). The attacker induces multiple requests containing varying plaintext, observing the length of the resulting ciphertext in each, which reveals a shorter length when a guess about the secret is correct.

+
+
+

Implementations communicating on a secure channel MUST NOT compress content that includes both confidential and attacker-controlled data unless separate compression dictionaries are used for each source of data. Compression MUST NOT be used if the source of data cannot be reliably determined. Generic stream compression, such as that provided by TLS MUST NOT be used with HTTP/2 (see Section 9.2).

+
+
+

Further considerations regarding the compression of header fields are described in + + [COMPRESSION] + .

+
+
+
+

10.7 Use of Padding

+
+

Padding within HTTP/2 is not intended as a replacement for general purpose padding, such as might be provided by TLS + [TLS12]. Redundant padding could even be counterproductive. Correct application can depend on having specific knowledge of the data that is being padded.

+
+
+

To mitigate attacks that rely on compression, disabling or limiting compression might be preferable to padding as a countermeasure.

+
+
+

Padding can be used to obscure the exact size of frame content, and is provided to mitigate specific attacks within HTTP. For example, attacks where compressed content includes both attacker-controlled plaintext and secret data (see for example, + + [BREACH] + ).

+
+
+

Use of padding can result in less protection than might seem immediately obvious. At best, padding only makes it more difficult for an attacker to infer length information by increasing the number of frames an attacker has to observe. Incorrectly implemented padding schemes can be easily defeated. In particular, randomized padding with a predictable distribution provides very little protection; similarly, padding payloads to a fixed size exposes information as payload sizes cross the fixed size boundary, which could be possible if an attacker can control plaintext.

+
+
+

Intermediaries SHOULD retain padding for DATA frames, but MAY drop padding for HEADERS and PUSH_PROMISE frames. A valid reason for an intermediary to change the amount of padding of frames is to improve the protections that padding provides.

+
+
+
+

10.8 Privacy Considerations +

+
+

Several characteristics of HTTP/2 provide an observer an opportunity to correlate actions of a single client or server over time. This includes the value of settings, the manner in which flow control windows are managed, the way priorities are allocated to streams, timing of reactions to stimulus, and handling of any features that are controlled by settings.

+
+
+

As far as this creates observable differences in behavior, they could be used as a basis for fingerprinting a specific client, as defined in Section 1.8 of + + [HTML5] + .

+
+
+

HTTP/2's preference for using a single TCP connection allows correlation of a user's activity on a site. If connections are reused for different origins, this allows tracking across those origins.

+
+
+

Because the PING and SETTINGS frames solicit immediate responses, they can be used by an endpoint to measure latency to their peer. This might have privacy implications in certain scenarios.

+
+
+
+
+

11. IANA Considerations

+
+

A string for identifying HTTP/2 is entered into the "Application Layer Protocol Negotiation (ALPN) Protocol IDs" registry established in + + [TLS-ALPN] + .

+
+
+

This document establishes a registry for frame types, settings, and error codes. These new registries are entered into a new "Hypertext Transfer Protocol (HTTP) 2 Parameters" section.

+
+
+

This document registers the + HTTP2-Settings header field for use in HTTP; and the 421 (Misdirected Request) status code.

+
+
+

This document registers the + PRI method for use in HTTP, to avoid collisions with the connection preface (Section 3.5).

+
+
+

11.1 Registration of HTTP/2 Identification Strings

+
+

This document creates two registrations for the identification of HTTP/2 in the "Application Layer Protocol Negotiation (ALPN) Protocol IDs" registry established in + + [TLS-ALPN] + .

+
+
+

The "h2" string identifies HTTP/2 when used over TLS:

+
+
Protocol:
+
HTTP/2 over TLS
+
Identification Sequence:
+
0x68 0x32 ("h2")
+
Specification:
+
This document
+
+
+
+

The "h2c" string identifies HTTP/2 when used over cleartext TCP:

+
+
Protocol:
+
HTTP/2 over TCP
+
Identification Sequence:
+
0x68 0x32 0x63 ("h2c")
+
Specification:
+
This document
+
+
+
+
+

11.2 Frame Type Registry

+
+

This document establishes a registry for HTTP/2 frame type codes. The "HTTP/2 Frame Type" registry manages an 8-bit space. The "HTTP/2 Frame Type" registry operates under either of the "IETF Review" or "IESG Approval" policies + [RFC5226] for values between 0x00 and 0xef, with values between 0xf0 and 0xff being reserved for experimental use.

+
+
+

New entries in this registry require the following information:

+
+
Frame Type:
+
A name or label for the frame type.
+
Code:
+
The 8-bit code assigned to the frame type.
+
Specification:
+
A reference to a specification that includes a description of the frame layout, its semantics, and flags that the frame type uses, including any parts of the frame that are conditionally present based on the value of flags.
+
+
+
+

The entries in the following table are registered by this document.

+
+
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
Frame TypeCodeSection
DATA0x0Section 6.1
HEADERS0x1Section 6.2
PRIORITY0x2Section 6.3
RST_STREAM0x3Section 6.4
SETTINGS0x4Section 6.5
PUSH_PROMISE0x5Section 6.6
PING0x6Section 6.7
GOAWAY0x7Section 6.8
WINDOW_UPDATE0x8Section 6.9
CONTINUATION0x9Section 6.10
+
+
+
+

11.3 Settings Registry

+
+

This document establishes a registry for HTTP/2 settings. The "HTTP/2 Settings" registry manages a 16-bit space. The "HTTP/2 Settings" registry operates under the "Expert Review" policy + [RFC5226] for values in the range from 0x0000 to 0xefff, with values between and 0xf000 and 0xffff being reserved for experimental use.

+
+
+

New registrations are advised to provide the following information:

+
+
Name:
+
A symbolic name for the setting. Specifying a setting name is optional.
+
Code:
+
The 16-bit code assigned to the setting.
+
Initial Value:
+
An initial value for the setting.
+
Specification:
+
An optional reference to a specification that describes the use of the setting.
+
+
+
+

An initial set of setting registrations can be found in Section 6.5.2.

+
+
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
NameCodeInitial ValueSpecification
HEADER_TABLE_SIZE0x14096Section 6.5.2
ENABLE_PUSH0x21Section 6.5.2
MAX_CONCURRENT_STREAMS0x3(infinite)Section 6.5.2
INITIAL_WINDOW_SIZE0x465535Section 6.5.2
MAX_FRAME_SIZE0x516384Section 6.5.2
MAX_HEADER_LIST_SIZE0x6(infinite)Section 6.5.2
+
+
+
+

11.4 Error Code Registry

+
+

This document establishes a registry for HTTP/2 error codes. The "HTTP/2 Error Code" registry manages a 32-bit space. The "HTTP/2 Error Code" registry operates under the "Expert Review" policy + [RFC5226].

+
+
+

Registrations for error codes are required to include a description of the error code. An expert reviewer is advised to examine new registrations for possible duplication with existing error codes. Use of existing registrations is to be encouraged, but not mandated.

+
+
+

New registrations are advised to provide the following information:

+
+
Name:
+
A name for the error code. Specifying an error code name is optional.
+
Code:
+
The 32-bit error code value.
+
Description:
+
A brief description of the error code semantics, longer if no detailed specification is provided.
+
Specification:
+
An optional reference for a specification that defines the error code.
+
+
+
+

The entries in the following table are registered by this document.

+
+
+
+

11.5 HTTP2-Settings Header Field Registration +

+
+

This section registers the + HTTP2-Settings header field in the Permanent Message Header Field Registry + [BCP90].

+
+
Header field name:
+
HTTP2-Settings
+
Applicable protocol:
+
http
+
Status:
+
standard
+
Author/Change controller:
+
IETF
+
Specification document(s):
+
Section 3.2.1 of this document
+
Related information:
+
This header field is only used by an HTTP/2 client for Upgrade-based negotiation.
+
+
+
+
+

11.6 PRI Method Registration +

+
+

This section registers the + PRI method in the HTTP Method Registry ( + + [RFC7231] + , Section 8.1).

+
+
Method Name:
+
PRI
+
Safe
+
Yes
+
Idempotent
+
Yes
+
Specification document(s)
+
Section 3.5 of this document
+
Related information:
+
This method is never used by an actual client. This method will appear to be used when an HTTP/1.1 server or intermediary attempts to parse an HTTP/2 connection preface.
+
+
+
+
+

11.7 The 421 (Misdirected Request) HTTP Status Code

+
+

This document registers the 421 (Misdirected Request) HTTP Status code in the Hypertext Transfer Protocol (HTTP) Status Code Registry ( + + [RFC7231] + , Section 8.2).

+
+
+
+
Status Code:
+
421
+
Short Description:
+
Misdirected Request
+
Specification:
+
Section 9.1.2 of this document
+
+
+
+
+
+

12. Acknowledgements +

+
+

This document includes substantial input from the following individuals:

+ +
+
+

13. References +

+

13.1 Normative References +

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
[COMPRESSION]Ruellan, H. and R. Peon, “HPACK - Header Compression for HTTP/2”, Internet-Draft draft-ietf-httpbis-header-compression-11 (work in progress), February 2015.
[COOKIE]Barth, A., “HTTP State Management Mechanism”, RFC 6265, April 2011, <http://www.rfc-editor.org/info/rfc6265>.
[FIPS186]NIST, “Digital Signature Standard (DSS)”, FIPS PUB 186-4, July 2013, <http://dx.doi.org/10.6028/NIST.FIPS.186-4>.
[RFC2119]Bradner, S., “Key words for use in RFCs to Indicate Requirement Levels”, BCP 14, RFC 2119, March 1997, <http://www.rfc-editor.org/info/rfc2119>.
[RFC2818]Rescorla, E., “HTTP Over TLS”, RFC 2818, May 2000, <http://www.rfc-editor.org/info/rfc2818>.
[RFC3986]Berners-Lee, T., Fielding, R., and L. Masinter, “Uniform Resource Identifier (URI): Generic Syntax”, STD 66, RFC 3986, January 2005, <http://www.rfc-editor.org/info/rfc3986>.
[RFC4648]Josefsson, S., “The Base16, Base32, and Base64 Data Encodings”, RFC 4648, October 2006, <http://www.rfc-editor.org/info/rfc4648>.
[RFC5226]Narten, T. and H. Alvestrand, “Guidelines for Writing an IANA Considerations Section in RFCs”, BCP 26, RFC 5226, May 2008, <http://www.rfc-editor.org/info/rfc5226>.
[RFC5234]Crocker, D. and P. Overell, “Augmented BNF for Syntax Specifications: ABNF”, STD 68, RFC 5234, January 2008, <http://www.rfc-editor.org/info/rfc5234>.
[RFC7230]Fielding, R., Ed. and J. Reschke, Ed., “Hypertext Transfer Protocol (HTTP/1.1): Message Syntax and Routing”, RFC 7230, June 2014, <http://www.rfc-editor.org/info/rfc7230>.
[RFC7231]Fielding, R., Ed. and J. Reschke, Ed., “Hypertext Transfer Protocol (HTTP/1.1): Semantics and Content”, RFC 7231, June 2014, <http://www.rfc-editor.org/info/rfc7231>.
[RFC7232]Fielding, R., Ed. and J. Reschke, Ed., “Hypertext Transfer Protocol (HTTP/1.1): Conditional Requests”, RFC 7232, June 2014, <http://www.rfc-editor.org/info/rfc7232>.
[RFC7233]Fielding, R., Ed., Lafon, Y., Ed., and J. Reschke, Ed., “Hypertext Transfer Protocol (HTTP/1.1): Range Requests”, RFC 7233, June 2014, <http://www.rfc-editor.org/info/rfc7233>.
[RFC7234]Fielding, R., Ed., Nottingham, M., Ed., and J. Reschke, Ed., “Hypertext Transfer Protocol (HTTP/1.1): Caching”, RFC 7234, June 2014, <http://www.rfc-editor.org/info/rfc7234>.
[RFC7235]Fielding, R., Ed. and J. Reschke, Ed., “Hypertext Transfer Protocol (HTTP/1.1): Authentication”, RFC 7235, June 2014, <http://www.rfc-editor.org/info/rfc7235>.
[TCP]Postel, J., “Transmission Control Protocol”, STD 7, RFC 793, September 1981, <http://www.rfc-editor.org/info/rfc793>.
[TLS-ALPN]Friedl, S., Popov, A., Langley, A., and E. Stephan, “Transport Layer Security (TLS) Application-Layer Protocol Negotiation Extension”, RFC 7301, July 2014, <http://www.rfc-editor.org/info/rfc7301>.
[TLS-ECDHE]Rescorla, E., “TLS Elliptic Curve Cipher Suites with SHA-256/384 and AES Galois Counter Mode (GCM)”, RFC 5289, August 2008, <http://www.rfc-editor.org/info/rfc5289>.
[TLS-EXT]Eastlake, D., “Transport Layer Security (TLS) Extensions: Extension Definitions”, RFC 6066, January 2011, <http://www.rfc-editor.org/info/rfc6066>.
[TLS12]Dierks, T. and E. Rescorla, “The Transport Layer Security (TLS) Protocol Version 1.2”, RFC 5246, August 2008, <http://www.rfc-editor.org/info/rfc5246>.
+

13.2 Informative References +

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
[ALT-SVC]Nottingham, M., McManus, P., and J. Reschke, “HTTP Alternative Services”, Internet-Draft draft-ietf-httpbis-alt-svc-06 (work in progress), February 2015.
[BCP90]Klyne, G., Nottingham, M., and J. Mogul, “Registration Procedures for Message Header Fields”, BCP 90, RFC 3864, September 2004, <http://www.rfc-editor.org/info/rfc3864>.
[BREACH]Gluck, Y., Harris, N., and A. Prado, “BREACH: Reviving the CRIME Attack”, July 2013, <http://breachattack.com/resources/BREACH%20-%20SSL,%20gone%20in%2030%20seconds.pdf>.
[HTML5]Hickson, I., Berjon, R., Faulkner, S., Leithead, T., Doyle Navara, E., O'Connor, E., and S. Pfeiffer, “HTML5”, W3C Recommendation REC-html5-20141028, October 2014, <http://www.w3.org/TR/2014/REC-html5-20141028/>. +
Latest version available at <http://www.w3.org/TR/html5/>.
[RFC3749]Hollenbeck, S., “Transport Layer Security Protocol Compression Methods”, RFC 3749, May 2004, <http://www.rfc-editor.org/info/rfc3749>.
[RFC4492]Blake-Wilson, S., Bolyard, N., Gupta, V., Hawk, C., and B. Moeller, “Elliptic Curve Cryptography (ECC) Cipher Suites for Transport Layer Security (TLS)”, RFC 4492, May 2006, <http://www.rfc-editor.org/info/rfc4492>.
[RFC6585]Nottingham, M. and R. Fielding, “Additional HTTP Status Codes”, RFC 6585, April 2012, <http://www.rfc-editor.org/info/rfc6585>.
[RFC7323]Borman, D., Braden, B., Jacobson, V., and R. Scheffenegger, “TCP Extensions for High Performance”, RFC 7323, September 2014, <http://www.rfc-editor.org/info/rfc7323>.
[TALKING]Huang, L-S., Chen, E., Barth, A., Rescorla, E., and C. Jackson, “Talking to Yourself for Fun and Profit”, 2011, <http://w2spconf.com/2011/papers/websocket.pdf>.
[TLSBCP]Sheffer, Y., Holz, R., and P. Saint-Andre, “Recommendations for Secure Use of TLS and DTLS”, Internet-Draft draft-ietf-uta-tls-bcp-08 (work in progress), December 2014.
+
+

A. TLS 1.2 Cipher Suite Black List

+
+

An HTTP/2 implementation MAY treat the negotiation of any of the following cipher suites with TLS 1.2 as a connection error (Section 5.4.1) of type INADEQUATE_SECURITY: TLS_NULL_WITH_NULL_NULL, TLS_RSA_WITH_NULL_MD5, TLS_RSA_WITH_NULL_SHA, TLS_RSA_EXPORT_WITH_RC4_40_MD5, TLS_RSA_WITH_RC4_128_MD5, TLS_RSA_WITH_RC4_128_SHA, TLS_RSA_EXPORT_WITH_RC2_CBC_40_MD5, TLS_RSA_WITH_IDEA_CBC_SHA, TLS_RSA_EXPORT_WITH_DES40_CBC_SHA, TLS_RSA_WITH_DES_CBC_SHA, TLS_RSA_WITH_3DES_EDE_CBC_SHA, TLS_DH_DSS_EXPORT_WITH_DES40_CBC_SHA, TLS_DH_DSS_WITH_DES_CBC_SHA, TLS_DH_DSS_WITH_3DES_EDE_CBC_SHA, TLS_DH_RSA_EXPORT_WITH_DES40_CBC_SHA, TLS_DH_RSA_WITH_DES_CBC_SHA, TLS_DH_RSA_WITH_3DES_EDE_CBC_SHA, TLS_DHE_DSS_EXPORT_WITH_DES40_CBC_SHA, TLS_DHE_DSS_WITH_DES_CBC_SHA, TLS_DHE_DSS_WITH_3DES_EDE_CBC_SHA, TLS_DHE_RSA_EXPORT_WITH_DES40_CBC_SHA, TLS_DHE_RSA_WITH_DES_CBC_SHA, TLS_DHE_RSA_WITH_3DES_EDE_CBC_SHA, TLS_DH_anon_EXPORT_WITH_RC4_40_MD5, TLS_DH_anon_WITH_RC4_128_MD5, TLS_DH_anon_EXPORT_WITH_DES40_CBC_SHA, TLS_DH_anon_WITH_DES_CBC_SHA, TLS_DH_anon_WITH_3DES_EDE_CBC_SHA, TLS_KRB5_WITH_DES_CBC_SHA, TLS_KRB5_WITH_3DES_EDE_CBC_SHA, TLS_KRB5_WITH_RC4_128_SHA, TLS_KRB5_WITH_IDEA_CBC_SHA, TLS_KRB5_WITH_DES_CBC_MD5, TLS_KRB5_WITH_3DES_EDE_CBC_MD5, TLS_KRB5_WITH_RC4_128_MD5, TLS_KRB5_WITH_IDEA_CBC_MD5, TLS_KRB5_EXPORT_WITH_DES_CBC_40_SHA, TLS_KRB5_EXPORT_WITH_RC2_CBC_40_SHA, TLS_KRB5_EXPORT_WITH_RC4_40_SHA, TLS_KRB5_EXPORT_WITH_DES_CBC_40_MD5, TLS_KRB5_EXPORT_WITH_RC2_CBC_40_MD5, TLS_KRB5_EXPORT_WITH_RC4_40_MD5, TLS_PSK_WITH_NULL_SHA, TLS_DHE_PSK_WITH_NULL_SHA, TLS_RSA_PSK_WITH_NULL_SHA, TLS_RSA_WITH_AES_128_CBC_SHA, TLS_DH_DSS_WITH_AES_128_CBC_SHA, TLS_DH_RSA_WITH_AES_128_CBC_SHA, TLS_DHE_DSS_WITH_AES_128_CBC_SHA, TLS_DHE_RSA_WITH_AES_128_CBC_SHA, TLS_DH_anon_WITH_AES_128_CBC_SHA, TLS_RSA_WITH_AES_256_CBC_SHA, TLS_DH_DSS_WITH_AES_256_CBC_SHA, TLS_DH_RSA_WITH_AES_256_CBC_SHA, TLS_DHE_DSS_WITH_AES_256_CBC_SHA, TLS_DHE_RSA_WITH_AES_256_CBC_SHA, TLS_DH_anon_WITH_AES_256_CBC_SHA, TLS_RSA_WITH_NULL_SHA256, TLS_RSA_WITH_AES_128_CBC_SHA256, TLS_RSA_WITH_AES_256_CBC_SHA256, TLS_DH_DSS_WITH_AES_128_CBC_SHA256, TLS_DH_RSA_WITH_AES_128_CBC_SHA256, TLS_DHE_DSS_WITH_AES_128_CBC_SHA256, TLS_RSA_WITH_CAMELLIA_128_CBC_SHA, TLS_DH_DSS_WITH_CAMELLIA_128_CBC_SHA, TLS_DH_RSA_WITH_CAMELLIA_128_CBC_SHA, TLS_DHE_DSS_WITH_CAMELLIA_128_CBC_SHA, TLS_DHE_RSA_WITH_CAMELLIA_128_CBC_SHA, TLS_DH_anon_WITH_CAMELLIA_128_CBC_SHA, TLS_DHE_RSA_WITH_AES_128_CBC_SHA256, TLS_DH_DSS_WITH_AES_256_CBC_SHA256, TLS_DH_RSA_WITH_AES_256_CBC_SHA256, TLS_DHE_DSS_WITH_AES_256_CBC_SHA256, TLS_DHE_RSA_WITH_AES_256_CBC_SHA256, TLS_DH_anon_WITH_AES_128_CBC_SHA256, TLS_DH_anon_WITH_AES_256_CBC_SHA256, TLS_RSA_WITH_CAMELLIA_256_CBC_SHA, TLS_DH_DSS_WITH_CAMELLIA_256_CBC_SHA, TLS_DH_RSA_WITH_CAMELLIA_256_CBC_SHA, TLS_DHE_DSS_WITH_CAMELLIA_256_CBC_SHA, TLS_DHE_RSA_WITH_CAMELLIA_256_CBC_SHA, TLS_DH_anon_WITH_CAMELLIA_256_CBC_SHA, TLS_PSK_WITH_RC4_128_SHA, TLS_PSK_WITH_3DES_EDE_CBC_SHA, TLS_PSK_WITH_AES_128_CBC_SHA, TLS_PSK_WITH_AES_256_CBC_SHA, TLS_DHE_PSK_WITH_RC4_128_SHA, TLS_DHE_PSK_WITH_3DES_EDE_CBC_SHA, TLS_DHE_PSK_WITH_AES_128_CBC_SHA, TLS_DHE_PSK_WITH_AES_256_CBC_SHA, TLS_RSA_PSK_WITH_RC4_128_SHA, TLS_RSA_PSK_WITH_3DES_EDE_CBC_SHA, TLS_RSA_PSK_WITH_AES_128_CBC_SHA, TLS_RSA_PSK_WITH_AES_256_CBC_SHA, TLS_RSA_WITH_SEED_CBC_SHA, TLS_DH_DSS_WITH_SEED_CBC_SHA, TLS_DH_RSA_WITH_SEED_CBC_SHA, TLS_DHE_DSS_WITH_SEED_CBC_SHA, TLS_DHE_RSA_WITH_SEED_CBC_SHA, TLS_DH_anon_WITH_SEED_CBC_SHA, TLS_RSA_WITH_AES_128_GCM_SHA256, TLS_RSA_WITH_AES_256_GCM_SHA384, TLS_DH_RSA_WITH_AES_128_GCM_SHA256, TLS_DH_RSA_WITH_AES_256_GCM_SHA384, TLS_DH_DSS_WITH_AES_128_GCM_SHA256, TLS_DH_DSS_WITH_AES_256_GCM_SHA384, TLS_DH_anon_WITH_AES_128_GCM_SHA256, TLS_DH_anon_WITH_AES_256_GCM_SHA384, TLS_PSK_WITH_AES_128_GCM_SHA256, TLS_PSK_WITH_AES_256_GCM_SHA384, TLS_RSA_PSK_WITH_AES_128_GCM_SHA256, TLS_RSA_PSK_WITH_AES_256_GCM_SHA384, TLS_PSK_WITH_AES_128_CBC_SHA256, TLS_PSK_WITH_AES_256_CBC_SHA384, TLS_PSK_WITH_NULL_SHA256, TLS_PSK_WITH_NULL_SHA384, TLS_DHE_PSK_WITH_AES_128_CBC_SHA256, TLS_DHE_PSK_WITH_AES_256_CBC_SHA384, TLS_DHE_PSK_WITH_NULL_SHA256, TLS_DHE_PSK_WITH_NULL_SHA384, TLS_RSA_PSK_WITH_AES_128_CBC_SHA256, TLS_RSA_PSK_WITH_AES_256_CBC_SHA384, TLS_RSA_PSK_WITH_NULL_SHA256, TLS_RSA_PSK_WITH_NULL_SHA384, TLS_RSA_WITH_CAMELLIA_128_CBC_SHA256, TLS_DH_DSS_WITH_CAMELLIA_128_CBC_SHA256, TLS_DH_RSA_WITH_CAMELLIA_128_CBC_SHA256, TLS_DHE_DSS_WITH_CAMELLIA_128_CBC_SHA256, TLS_DHE_RSA_WITH_CAMELLIA_128_CBC_SHA256, TLS_DH_anon_WITH_CAMELLIA_128_CBC_SHA256, TLS_RSA_WITH_CAMELLIA_256_CBC_SHA256, TLS_DH_DSS_WITH_CAMELLIA_256_CBC_SHA256, TLS_DH_RSA_WITH_CAMELLIA_256_CBC_SHA256, TLS_DHE_DSS_WITH_CAMELLIA_256_CBC_SHA256, TLS_DHE_RSA_WITH_CAMELLIA_256_CBC_SHA256, TLS_DH_anon_WITH_CAMELLIA_256_CBC_SHA256, TLS_EMPTY_RENEGOTIATION_INFO_SCSV, TLS_ECDH_ECDSA_WITH_NULL_SHA, TLS_ECDH_ECDSA_WITH_RC4_128_SHA, TLS_ECDH_ECDSA_WITH_3DES_EDE_CBC_SHA, TLS_ECDH_ECDSA_WITH_AES_128_CBC_SHA, TLS_ECDH_ECDSA_WITH_AES_256_CBC_SHA, TLS_ECDHE_ECDSA_WITH_NULL_SHA, TLS_ECDHE_ECDSA_WITH_RC4_128_SHA, TLS_ECDHE_ECDSA_WITH_3DES_EDE_CBC_SHA, TLS_ECDHE_ECDSA_WITH_AES_128_CBC_SHA, TLS_ECDHE_ECDSA_WITH_AES_256_CBC_SHA, TLS_ECDH_RSA_WITH_NULL_SHA, TLS_ECDH_RSA_WITH_RC4_128_SHA, TLS_ECDH_RSA_WITH_3DES_EDE_CBC_SHA, TLS_ECDH_RSA_WITH_AES_128_CBC_SHA, TLS_ECDH_RSA_WITH_AES_256_CBC_SHA, TLS_ECDHE_RSA_WITH_NULL_SHA, TLS_ECDHE_RSA_WITH_RC4_128_SHA, TLS_ECDHE_RSA_WITH_3DES_EDE_CBC_SHA, TLS_ECDHE_RSA_WITH_AES_128_CBC_SHA, TLS_ECDHE_RSA_WITH_AES_256_CBC_SHA, TLS_ECDH_anon_WITH_NULL_SHA, TLS_ECDH_anon_WITH_RC4_128_SHA, TLS_ECDH_anon_WITH_3DES_EDE_CBC_SHA, TLS_ECDH_anon_WITH_AES_128_CBC_SHA, TLS_ECDH_anon_WITH_AES_256_CBC_SHA, TLS_SRP_SHA_WITH_3DES_EDE_CBC_SHA, TLS_SRP_SHA_RSA_WITH_3DES_EDE_CBC_SHA, TLS_SRP_SHA_DSS_WITH_3DES_EDE_CBC_SHA, TLS_SRP_SHA_WITH_AES_128_CBC_SHA, TLS_SRP_SHA_RSA_WITH_AES_128_CBC_SHA, TLS_SRP_SHA_DSS_WITH_AES_128_CBC_SHA, TLS_SRP_SHA_WITH_AES_256_CBC_SHA, TLS_SRP_SHA_RSA_WITH_AES_256_CBC_SHA, TLS_SRP_SHA_DSS_WITH_AES_256_CBC_SHA, TLS_ECDHE_ECDSA_WITH_AES_128_CBC_SHA256, TLS_ECDHE_ECDSA_WITH_AES_256_CBC_SHA384, TLS_ECDH_ECDSA_WITH_AES_128_CBC_SHA256, TLS_ECDH_ECDSA_WITH_AES_256_CBC_SHA384, TLS_ECDHE_RSA_WITH_AES_128_CBC_SHA256, TLS_ECDHE_RSA_WITH_AES_256_CBC_SHA384, TLS_ECDH_RSA_WITH_AES_128_CBC_SHA256, TLS_ECDH_RSA_WITH_AES_256_CBC_SHA384, TLS_ECDH_ECDSA_WITH_AES_128_GCM_SHA256, TLS_ECDH_ECDSA_WITH_AES_256_GCM_SHA384, TLS_ECDH_RSA_WITH_AES_128_GCM_SHA256, TLS_ECDH_RSA_WITH_AES_256_GCM_SHA384, TLS_ECDHE_PSK_WITH_RC4_128_SHA, TLS_ECDHE_PSK_WITH_3DES_EDE_CBC_SHA, TLS_ECDHE_PSK_WITH_AES_128_CBC_SHA, TLS_ECDHE_PSK_WITH_AES_256_CBC_SHA, TLS_ECDHE_PSK_WITH_AES_128_CBC_SHA256, TLS_ECDHE_PSK_WITH_AES_256_CBC_SHA384, TLS_ECDHE_PSK_WITH_NULL_SHA, TLS_ECDHE_PSK_WITH_NULL_SHA256, TLS_ECDHE_PSK_WITH_NULL_SHA384, TLS_RSA_WITH_ARIA_128_CBC_SHA256, TLS_RSA_WITH_ARIA_256_CBC_SHA384, TLS_DH_DSS_WITH_ARIA_128_CBC_SHA256, TLS_DH_DSS_WITH_ARIA_256_CBC_SHA384, TLS_DH_RSA_WITH_ARIA_128_CBC_SHA256, TLS_DH_RSA_WITH_ARIA_256_CBC_SHA384, TLS_DHE_DSS_WITH_ARIA_128_CBC_SHA256, TLS_DHE_DSS_WITH_ARIA_256_CBC_SHA384, TLS_DHE_RSA_WITH_ARIA_128_CBC_SHA256, TLS_DHE_RSA_WITH_ARIA_256_CBC_SHA384, TLS_DH_anon_WITH_ARIA_128_CBC_SHA256, TLS_DH_anon_WITH_ARIA_256_CBC_SHA384, TLS_ECDHE_ECDSA_WITH_ARIA_128_CBC_SHA256, TLS_ECDHE_ECDSA_WITH_ARIA_256_CBC_SHA384, TLS_ECDH_ECDSA_WITH_ARIA_128_CBC_SHA256, TLS_ECDH_ECDSA_WITH_ARIA_256_CBC_SHA384, TLS_ECDHE_RSA_WITH_ARIA_128_CBC_SHA256, TLS_ECDHE_RSA_WITH_ARIA_256_CBC_SHA384, TLS_ECDH_RSA_WITH_ARIA_128_CBC_SHA256, TLS_ECDH_RSA_WITH_ARIA_256_CBC_SHA384, TLS_RSA_WITH_ARIA_128_GCM_SHA256, TLS_RSA_WITH_ARIA_256_GCM_SHA384, TLS_DH_RSA_WITH_ARIA_128_GCM_SHA256, TLS_DH_RSA_WITH_ARIA_256_GCM_SHA384, TLS_DH_DSS_WITH_ARIA_128_GCM_SHA256, TLS_DH_DSS_WITH_ARIA_256_GCM_SHA384, TLS_DH_anon_WITH_ARIA_128_GCM_SHA256, TLS_DH_anon_WITH_ARIA_256_GCM_SHA384, TLS_ECDH_ECDSA_WITH_ARIA_128_GCM_SHA256, TLS_ECDH_ECDSA_WITH_ARIA_256_GCM_SHA384, TLS_ECDH_RSA_WITH_ARIA_128_GCM_SHA256, TLS_ECDH_RSA_WITH_ARIA_256_GCM_SHA384, TLS_PSK_WITH_ARIA_128_CBC_SHA256, TLS_PSK_WITH_ARIA_256_CBC_SHA384, TLS_DHE_PSK_WITH_ARIA_128_CBC_SHA256, TLS_DHE_PSK_WITH_ARIA_256_CBC_SHA384, TLS_RSA_PSK_WITH_ARIA_128_CBC_SHA256, TLS_RSA_PSK_WITH_ARIA_256_CBC_SHA384, TLS_PSK_WITH_ARIA_128_GCM_SHA256, TLS_PSK_WITH_ARIA_256_GCM_SHA384, TLS_RSA_PSK_WITH_ARIA_128_GCM_SHA256, TLS_RSA_PSK_WITH_ARIA_256_GCM_SHA384, TLS_ECDHE_PSK_WITH_ARIA_128_CBC_SHA256, TLS_ECDHE_PSK_WITH_ARIA_256_CBC_SHA384, TLS_ECDHE_ECDSA_WITH_CAMELLIA_128_CBC_SHA256, TLS_ECDHE_ECDSA_WITH_CAMELLIA_256_CBC_SHA384, TLS_ECDH_ECDSA_WITH_CAMELLIA_128_CBC_SHA256, TLS_ECDH_ECDSA_WITH_CAMELLIA_256_CBC_SHA384, TLS_ECDHE_RSA_WITH_CAMELLIA_128_CBC_SHA256, TLS_ECDHE_RSA_WITH_CAMELLIA_256_CBC_SHA384, TLS_ECDH_RSA_WITH_CAMELLIA_128_CBC_SHA256, TLS_ECDH_RSA_WITH_CAMELLIA_256_CBC_SHA384, TLS_RSA_WITH_CAMELLIA_128_GCM_SHA256, TLS_RSA_WITH_CAMELLIA_256_GCM_SHA384, TLS_DH_RSA_WITH_CAMELLIA_128_GCM_SHA256, TLS_DH_RSA_WITH_CAMELLIA_256_GCM_SHA384, TLS_DH_DSS_WITH_CAMELLIA_128_GCM_SHA256, TLS_DH_DSS_WITH_CAMELLIA_256_GCM_SHA384, TLS_DH_anon_WITH_CAMELLIA_128_GCM_SHA256, TLS_DH_anon_WITH_CAMELLIA_256_GCM_SHA384, TLS_ECDH_ECDSA_WITH_CAMELLIA_128_GCM_SHA256, TLS_ECDH_ECDSA_WITH_CAMELLIA_256_GCM_SHA384, TLS_ECDH_RSA_WITH_CAMELLIA_128_GCM_SHA256, TLS_ECDH_RSA_WITH_CAMELLIA_256_GCM_SHA384, TLS_PSK_WITH_CAMELLIA_128_GCM_SHA256, TLS_PSK_WITH_CAMELLIA_256_GCM_SHA384, TLS_RSA_PSK_WITH_CAMELLIA_128_GCM_SHA256, TLS_RSA_PSK_WITH_CAMELLIA_256_GCM_SHA384, TLS_PSK_WITH_CAMELLIA_128_CBC_SHA256, TLS_PSK_WITH_CAMELLIA_256_CBC_SHA384, TLS_DHE_PSK_WITH_CAMELLIA_128_CBC_SHA256, TLS_DHE_PSK_WITH_CAMELLIA_256_CBC_SHA384, TLS_RSA_PSK_WITH_CAMELLIA_128_CBC_SHA256, TLS_RSA_PSK_WITH_CAMELLIA_256_CBC_SHA384, TLS_ECDHE_PSK_WITH_CAMELLIA_128_CBC_SHA256, TLS_ECDHE_PSK_WITH_CAMELLIA_256_CBC_SHA384, TLS_RSA_WITH_AES_128_CCM, TLS_RSA_WITH_AES_256_CCM, TLS_RSA_WITH_AES_128_CCM_8, TLS_RSA_WITH_AES_256_CCM_8, TLS_PSK_WITH_AES_128_CCM, TLS_PSK_WITH_AES_256_CCM, TLS_PSK_WITH_AES_128_CCM_8, TLS_PSK_WITH_AES_256_CCM_8.

+
+
Note:
+
This list was assembled from the set of registered TLS cipher suites at the time of writing. This list includes those cipher suites that do not offer an ephemeral key exchange and those that are based on the TLS null, stream or block cipher type (as defined in Section 6.2.3 of + + [TLS12] + ). Additional cipher suites with these properties could be defined; these would not be explicitly prohibited.
+
+
+
+
+

B. Change Log

+
+

This section is to be removed by RFC Editor before publication.

+
+
+

B.1 Since draft-ietf-httpbis-http2-15

+
+

Enabled the sending of PRIORITY for any stream state.

+
+
+

Added a cipher suite blacklist and made several changes to the TLS usage section.

+
+
+
+

B.2 Since draft-ietf-httpbis-http2-14

+
+

Renamed Not Authoritative status code to Misdirected Request.

+
+
+

Added HTTP_1_1_REQUIRED error code.

+
+
+
+

B.3 Since draft-ietf-httpbis-http2-13

+
+

Pseudo-header fields are now required to appear strictly before regular ones.

+
+
+

Restored 1xx series status codes, except 101.

+
+
+

Changed frame length field 24-bits. Expanded frame header to 9 octets. Added a setting to limit the damage.

+
+
+

Added a setting to advise peers of header set size limits.

+
+
+

Removed segments.

+
+
+

Made non-semantic-bearing HEADERS frames illegal in the HTTP mapping.

+
+
+
+

B.4 Since draft-ietf-httpbis-http2-12

+
+

Restored extensibility options.

+
+
+

Restricting TLS cipher suites to AEAD only.

+
+
+

Removing Content-Encoding requirements.

+
+
+

Permitting the use of PRIORITY after stream close.

+
+
+

Removed ALTSVC frame.

+
+
+

Removed BLOCKED frame.

+
+
+

Reducing the maximum padding size to 256 octets; removing padding from CONTINUATION frames.

+
+
+

Removed per-frame GZIP compression.

+
+
+
+

B.5 Since draft-ietf-httpbis-http2-11

+
+

Added BLOCKED frame (at risk).

+
+
+

Simplified priority scheme.

+
+
+

Added DATA per-frame GZIP compression.

+
+
+
+

B.6 Since draft-ietf-httpbis-http2-10

+
+

Changed "connection header" to "connection preface" to avoid confusion.

+
+
+

Added dependency-based stream prioritization.

+
+
+

Added "h2c" identifier to distinguish between cleartext and secured HTTP/2.

+
+
+

Adding missing padding to PUSH_PROMISE.

+
+
+

Integrate ALTSVC frame and supporting text.

+
+
+

Dropping requirement on "deflate" Content-Encoding.

+
+
+

Improving security considerations around use of compression.

+
+
+
+

B.7 Since draft-ietf-httpbis-http2-09

+
+

Adding padding for data frames.

+
+
+

Renumbering frame types, error codes, and settings.

+
+
+

Adding INADEQUATE_SECURITY error code.

+
+
+

Updating TLS usage requirements to 1.2; forbidding TLS compression.

+
+
+

Removing extensibility for frames and settings.

+
+
+

Changing setting identifier size.

+
+
+

Removing the ability to disable flow control.

+
+
+

Changing the protocol identification token to "h2".

+
+
+

Changing the use of :authority to make it optional and to allow userinfo in non-HTTP cases.

+
+
+

Allowing split on 0x0 for Cookie.

+
+
+

Reserved PRI method in HTTP/1.1 to avoid possible future collisions.

+
+
+
+

B.8 Since draft-ietf-httpbis-http2-08

+
+

Added cookie crumbling for more efficient header compression.

+
+
+

Added header field ordering with the value-concatenation mechanism.

+
+
+
+

B.9 Since draft-ietf-httpbis-http2-07

+
+

Marked draft for implementation.

+
+
+
+

B.10 Since draft-ietf-httpbis-http2-06

+
+

Adding definition for CONNECT method.

+
+
+

Constraining the use of push to safe, cacheable methods with no request body.

+
+
+

Changing from :host to :authority to remove any potential confusion.

+
+
+

Adding setting for header compression table size.

+
+
+

Adding settings acknowledgement.

+
+
+

Removing unnecessary and potentially problematic flags from CONTINUATION.

+
+
+

Added denial of service considerations.

+
+
+
+

B.11 Since draft-ietf-httpbis-http2-05

+
+

Marking the draft ready for implementation.

+
+
+

Renumbering END_PUSH_PROMISE flag.

+
+
+

Editorial clarifications and changes.

+
+
+
+

B.12 Since draft-ietf-httpbis-http2-04

+
+

Added CONTINUATION frame for HEADERS and PUSH_PROMISE.

+
+
+

PUSH_PROMISE is no longer implicitly prohibited if SETTINGS_MAX_CONCURRENT_STREAMS is zero.

+
+
+

Push expanded to allow all safe methods without a request body.

+
+
+

Clarified the use of HTTP header fields in requests and responses. Prohibited HTTP/1.1 hop-by-hop header fields.

+
+
+

Requiring that intermediaries not forward requests with missing or illegal routing :-headers.

+
+
+

Clarified requirements around handling different frames after stream close, stream reset and GOAWAY.

+
+
+

Added more specific prohibitions for sending of different frame types in various stream states.

+
+
+

Making the last received setting value the effective value.

+
+
+

Clarified requirements on TLS version, extension and ciphers.

+
+
+
+

B.13 Since draft-ietf-httpbis-http2-03

+
+

Committed major restructuring atrocities.

+
+
+

Added reference to first header compression draft.

+
+
+

Added more formal description of frame lifecycle.

+
+
+

Moved END_STREAM (renamed from FINAL) back to HEADERS/DATA.

+
+
+

Removed HEADERS+PRIORITY, added optional priority to HEADERS frame.

+
+
+

Added PRIORITY frame.

+
+
+
+

B.14 Since draft-ietf-httpbis-http2-02

+
+

Added continuations to frames carrying header blocks.

+
+
+

Replaced use of "session" with "connection" to avoid confusion with other HTTP stateful concepts, like cookies.

+
+
+

Removed "message".

+
+
+

Switched to TLS ALPN from NPN.

+
+
+

Editorial changes.

+
+
+
+

B.15 Since draft-ietf-httpbis-http2-01

+
+

Added IANA considerations section for frame types, error codes and settings.

+
+
+

Removed data frame compression.

+
+
+

Added PUSH_PROMISE.

+
+
+

Added globally applicable flags to framing.

+
+
+

Removed zlib-based header compression mechanism.

+
+
+

Updated references.

+
+
+

Clarified stream identifier reuse.

+
+
+

Removed CREDENTIALS frame and associated mechanisms.

+
+
+

Added advice against naive implementation of flow control.

+
+
+

Added session header section.

+
+
+

Restructured frame header. Removed distinction between data and control frames.

+
+
+

Altered flow control properties to include session-level limits.

+
+
+

Added note on cacheability of pushed resources and multiple tenant servers.

+
+
+

Changed protocol label form based on discussions.

+
+
+
+

B.16 Since draft-ietf-httpbis-http2-00

+
+

Changed title throughout.

+
+
+

Removed section on Incompatibilities with SPDY draft#2.

+
+ +
+

Replaced abstract and introduction.

+
+
+

Added section on starting HTTP/2.0, including upgrade mechanism.

+
+
+

Removed unused references.

+
+ +
+
+

B.17 Since draft-mbelshe-httpbis-spdy-00

+
+

Adopted as base for draft-ietf-httpbis-http2.

+
+
+

Updated authors/editors list.

+
+
+

Added status note.

+
+
+
+
+

Authors' Addresses

+

Mike Belshe +
Twist +
Email: mbelshe@chromium.org

+

Roberto Peon +
Google, Inc +
Email: fenix@google.com

+

Martin Thomson (editor) +
Mozilla +
331 E Evelyn Street +
Mountain View, CA 94041 +
US +
Email: martin.thomson@gmail.com

+
+
feedback
+ + + diff --git a/test/test-pages/keep-images/expected.html b/test/test-pages/keep-images/expected.html index df2bbd0..e348f63 100644 --- a/test/test-pages/keep-images/expected.html +++ b/test/test-pages/keep-images/expected.html @@ -1,160 +1,156 @@
-
-
-
-
-
-
-
-
-
-

Welcome to DoctorX’s Barcelona lab, where the drugs you bought online are tested for safety and purity. No questions asked.

-
-
-
-

Standing at a table in a chemistry lab in Barcelona, Cristina Gil Lladanosa tears open a silver, smell-proof protective envelope. She slides out a transparent bag full of crystals. Around her, machines whir and hum, and other researchers mill around in long, white coats.

-

She is holding the lab’s latest delivery of a drug bought from the “deep web,” the clandestine corner of the internet that isn’t reachable by normal search engines, and is home to some sites that require special software to access. Labeled as MDMA (the street term is ecstasy), this sample has been shipped from Canada. Lladanosa and her colleague Iván Fornís Espinosa have also received drugs, anonymously, from people in China, Australia, Europe and the United States.

-

“Here we have speed, MDMA, cocaine, pills,” Lladanosa says, pointing to vials full of red, green, blue and clear solutions sitting in labeled boxes.

-
-
-
-
-
Cristina Gil Lladanosa, at the Barcelona testing lab | photo by Joan Bardeletti
-
-
-
-

Since 2011, with the launch of Silk Road, anybody has been able to safely buy illegal drugs from the deep web and have them delivered to their door. Though the FBI shut down that black market in October 2013, other outlets have emerged to fill its role. For the last 10 months the lab at which Lladanosa and Espinosa work has offered a paid testing service of those drugs. By sending in samples for analysis, users can know exactly what it is they are buying, and make a more informed decision about whether to ingest the substance. The group, called Energy Control, which has being running “harm reduction” programs since 1999, is the first to run a testing service explicitly geared towards verifying those purchases from the deep web.

-

Before joining Energy Control, Lladanosa briefly worked at a pharmacy, whereas Espinosa spent 14 years doing drug analysis. Working at Energy Control is “more gratifying,” and “rewarding” than her previous jobs, Lladanosa told me. They also receive help from a group of volunteers, made up of a mixture of “squatters,” as Espinosa put it, and medical students, who prepare the samples for testing.

-

After weighing out the crystals, aggressively mixing it with methanol until dissolved, and delicately pouring the liquid into a tiny brown bottle, Lladanosa, a petite woman who is nearly engulfed by her lab coat, is now ready to test the sample. She loads a series of three trays on top of a large white appliance sitting on a table, called a gas chromatograph (GC). A jungle of thick pipes hang from the lab’s ceiling behind it.

-
-
-
-
-
Photo by Joan Bardeletti
-
-
-
-

“Chromatography separates all the substances,” Lladanosa says as she loads the machine with an array of drugs sent from the deep web and local Spanish users. It can tell whether a sample is pure or contaminated, and if the latter, with what.

-

Rushes of hot air blow across the desk as the gas chromatograph blasts the sample at 280 degrees Celsius. Thirty minutes later the machine’s robotic arm automatically moves over to grip another bottle. The machine will continue cranking through the 150 samples in the trays for most of the work week.

-
-
-
-
-
Photo by Joan Bardeletti
-
-
-
-

To get the drugs to Barcelona, a user mails at least 10 milligrams of a substance to the offices of the Asociación Bienestar y Desarrollo, the non-government organization that oversees Energy Control. The sample then gets delivered to the testing service’s laboratory, at the Barcelona Biomedical Research Park, a futuristic, seven story building sitting metres away from the beach. Energy Control borrows its lab space from a biomedical research group for free.

-

The tests cost 50 Euro per sample. Users pay, not surprisingly, with Bitcoin. In the post announcing Energy Control’s service on the deep web, the group promised that “All profits of this service are set aside of maintenance of this project.”

-

About a week after testing, those results are sent in a PDF to an email address provided by the anonymous client.

-

“The process is quite boring, because you are in a routine,” Lladanosa says. But one part of the process is consistently surprising: that moment when the results pop up on the screen. “Every time it’s something different.” For instance, one cocaine sample she had tested also contained phenacetin, a painkiller added to increase the product’s weight; lidocaine, an anesthetic that numbs the gums, giving the impression that the user is taking higher quality cocaine; and common caffeine.

-
-
-
-

The deep web drug lab is the brainchild of Fernando Caudevilla, a Spanish physician who is better known as “DoctorX” on the deep web, a nickname given to him by his Energy Control co-workers because of his earlier writing about the history, risks and recreational culture of MDMA. In the physical world, Caudevilla has worked for over a decade with Energy Control on various harm reduction focused projects, most of which have involved giving Spanish illegal drug users medical guidance, and often writing leaflets about the harms of certain substances.

-
-
-
-
-
Fernando Caudevilla, AKA DoctorX. Photo: Joseph Cox
-
-
-
-

Caudevilla first ventured into Silk Road forums in April 2013. “I would like to contribute to this forum offering professional advice in topics related to drug use and health,” he wrote in an introductory post, using his DoctorX alias. Caudevilla offered to provide answers to questions that a typical doctor is not prepared, or willing, to respond to, at least not without a lecture or a judgment. “This advice cannot replace a complete face-to-face medical evaluation,” he wrote, “but I know how difficult it can be to talk frankly about these things.”

-

The requests flooded in. A diabetic asked what effect MDMA has on blood sugar; another what the risks of frequent psychedelic use were for a young person. Someone wanted to know whether amphetamine use should be avoided during lactation. In all, Fernando’s thread received over 50,000 visits and 300 questions before the FBI shut down Silk Road.

-

“He’s amazing. A gift to this community,” one user wrote on the Silk Road 2.0 forum, a site that sprang up after the original. “His knowledge is invaluable, and never comes with any judgment.” Up until recently, Caudevilla answered questions on the marketplace “Evolution.” Last week, however, the administrators of that site pulled a scam, shutting the market down and escaping with an estimated $12 million worth of Bitcoin.

-

Caudevilla’s transition from dispensing advice to starting up a no-questions-asked drug testing service came as a consequence of his experience on the deep web. He’d wondered whether he could help bring more harm reduction services to a marketplace without controls. The Energy Control project, as part of its mandate of educating drug users and preventing harm, had already been carrying out drug testing for local Spanish users since 2001, at music festivals, night clubs, or through a drop-in service at a lab in Madrid.

-

“I thought, we are doing this in Spain, why don’t we do an international drug testing service?” Caudevilla told me when I visited the other Energy Control lab, in Madrid. Caudevilla, a stocky character with ear piercings and short, shaved hair, has eyes that light up whenever he discusses the world of the deep web. Later, via email, he elaborated that it was not a hard sell. “It was not too hard to convince them,” he wrote me. Clearly, Energy Control believed that the reputation he had earned as an unbiased medical professional on the deep web might carry over to the drug analysis service, where one needs to establish “credibility, trustworthiness, [and] transparency,” Caudevilla said. “We could not make mistakes,” he added.

-
-
-
-
-
Photo: Joseph Cox
-
-
-
-
-
-
-

While the Energy Control lab in Madrid lab only tests Spanish drugs from various sources, it is the Barcelona location which vets the substances bought in the shadowy recesses of of the deep web. Caudevilla no longer runs it, having handed it over to his colleague Ana Muñoz. She maintains a presence on the deep web forums, answers questions from potential users, and sends back reports when they are ready.

-

The testing program exists in a legal grey area. The people who own the Barcelona lab are accredited to experiment with and handle drugs, but Energy Control doesn’t have this permission itself, at least not in writing.

-

“We have a verbal agreement with the police and other authorities. They already know what we are doing,” Lladanosa tells me. It is a pact of mutual benefit. Energy Control provides the police with information on batches of drugs in Spain, whether they’re from the deep web or not, Espinosa says. They also contribute to the European Monitoring Centre for Drugs and Drug Addiction’s early warning system, a collaboration that attempts to spread information about dangerous drugs as quickly as possible.

-

By the time of my visit in February, Energy Control had received over 150 samples from the deep web and have been receiving more at a rate of between 4 and 8 a week. Traditional drugs, such as cocaine and MDMA, make up about 70 percent of the samples tested, but the Barcelona lab has also received samples of the prescription pill codeine, research chemicals and synthetic cannabinoids, and even pills of Viagra.

-
-
-
-
-
Photo by Joan Bardeletti
-
-
-
-

So it’s fair to make a tentative judgement on what people are paying for on the deep web. The verdict thus far? Overall, drugs on the deep web appear to be of much higher quality than those found on the street.

-

“In general, the cocaine is amazing,” says Caudevilla, saying that the samples they’ve seen have purities climbing towards 80 or 90 percent, and some even higher. To get an idea of how unusual this is, take a look at the United Nations Office on Drugs and Crime World Drug Report 2014, which reports that the average quality of street cocaine in Spain is just over 40 percent, while in the United Kingdom it is closer to 30 percent.“We have found 100 percent [pure] cocaine,” he adds. “That’s really, really strange. That means that, technically, this cocaine has been purified, with clandestine methods.”

-

Naturally, identifying vendors who sell this top-of-the-range stuff is one of the reasons that people have sent samples to Energy Control. Caudevilla was keen to stress that, officially, Energy Control’s service “is not intended to be a control of drug quality,” meaning a vetting process for identifying the best sellers, but that is exactly how some people have been using it.

-

As one buyer on the Evolution market, elmo666, wrote to me over the site’s messaging system, “My initial motivations were selfish. My primary motivation was to ensure that I was receiving and continue to receive a high quality product, essentially to keep the vendor honest as far as my interactions with them went.”

-

Vendors on deep web markets advertise their product just like any other outlet does, using flash sales, gimmicky giveaways and promises of drugs that are superior to those of their competitors. The claims, however, can turn out to be empty: despite the test results that show that deep web cocaine vendors typically sell product that is of a better quality than that found on the street, in plenty of cases, the drugs are nowhere near as pure as advertised.

-

“You won’t be getting anything CLOSE to what you paid for,” one user complained about the cocaine from ‘Mirkov’, a vendor on Evolution. “He sells 65% not 95%.”

-
-
-
-
-
Photo by Joan Bardeletti
-
-
-
-
-
-
-

Despite the prevalence of people using the service to gauge the quality of what goes up their nose, many users send samples to Energy Control in the spirit of its original mission: keeping themselves alive and healthy. The worst case scenario from drugs purchased on the deep web is, well the worst case. That was the outcome when Patrick McMullen, a 17-year-old Scottish student, ingested half a gram of MDMA and three tabs of LSD, reportedly purchased from the Silk Road. While talking to his friends on Skype, his words became slurred and he passed out. Paramedics could not revive him. The coroner for that case, Sherrif Payne, who deemed the cause of death ecstasy toxicity, told The Independent “You never know the purity of what you are taking and you can easily come unstuck.”

-

ScreamMyName, a deep web user who has been active since the original Silk Road, wants to alert users to the dangerous chemicals that are often mixed with drugs, and is using Energy Control as a means to do so.

-

“We’re at a time where some vendors are outright sending people poison. Some do it unknowingly,” ScreamMyName told me in an encrypted message. “Cocaine production in South America is often tainted with either levamisole or phenacetine. Both poison to humans and both with severe side effects.”

-

In the case of Levamisole, those prescribing it are often not doctors but veterinarians, as Levamisole is commonly used on animals, primarily for the treatment of worms. If ingested by humans it can lead to cases of extreme eruptions of the skin, as documented in a study from researchers at the University of California, San Francisco. But Lladanosa has found Levamisole in cocaine samples; dealers use it to increase the product weight, allowing them to stretch their batch further for greater profit — and also, she says, because Levamisole has a strong stimulant effect.

-

“It got me sick as fuck,” Dr. Feel, an Evolution user, wrote on the site’s forums after consuming cocaine that had been cut with 23 percent Levamisole, and later tested by Energy Control. “I was laid up in bed for several days because of that shit. The first night I did it, I thought I was going to die. I nearly drove myself to the ER.”

-

“More people die because of tainted drugs than the drugs themselves,” Dr. Feel added. “It’s the cuts and adulterants that are making people sick and killing them.”

-
-
-
-
-
Photo by Joan Bardeletti
-
-
-
-

The particular case of cocaine cut with Levamisole is one of the reasons that ScreamMyName has been pushing for more drug testing on the deep web markets. “I recognize that drug use isn’t exactly healthy, but why exacerbate the problem?” he told me when I contacted him after his post. “[Energy Control] provides a way for users to test the drugs they’ll use and for these very users to know what it is they’re putting in their bodies. Such services are in very short supply.”

-

After sending a number of Energy Control tests himself, ScreamMyName started a de facto crowd-sourcing campaign to get more drugs sent to the lab, and then shared the results, after throwing in some cash to get the ball rolling. He set up a Bitcoin wallet, with the hope that users might chip in to fund further tests. At the time of writing, the wallet has received a total of 1.81 bitcoins; around $430 at today’s exchange rates.

-

In posts to the Evolution community, ScreamMyName pitched this project as something that will benefit users and keep drug dealer honest. “When the funds build up to a point where we can purchase an [Energy Control] test fee, we’ll do a US thread poll for a few days and try to cohesively decide on what vendor to test,” he continued.

-
-
-
-
-
Photo by Joan Bardeletti
-
-
-
-

Other members of the community have been helping out, too. PlutoPete, a vendor from the original Silk Road who sold cannabis seeds and other legal items, has provided ScreamMyName with packaging to safely send the samples to Barcelona. “A box of baggies, and a load of different moisture barrier bags,” PlutoPete told me over the phone. “That’s what all the vendors use.”

-

It’s a modest program so far. ScreamMyName told me that so far he had gotten enough public funding to purchase five different Energy Control tests, in addition to the ten or so he’s sent himself so far. “The program created is still in its infancy and it is growing and changing as we go along but I have a lot of faith in what we’re doing,” he says.

-

But the spirit is contagious: elmo666, the other deep web user testing cocaine, originally kept the results of the drug tests to himself, but he, too, saw a benefit to distributing the data. “It is clear that it is a useful service to other users, keeping vendors honest and drugs (and their users) safe,” he told me. He started to report his findings to others on the forums, and then created a thread with summaries of the test results, as well as comments from the vendors if they provided it. Other users were soon basing their decisions on what to buy on elmo666‘s tests.

-

“I’m defo trying the cola based on the incredibly helpful elmo and his energy control results and recommendations,” wrote user jayk1984. On top of this, elmo666 plans to launch an independent site on the deep web that will collate all of these results, which should act as a resource for users of all the marketplaces.

-

As word of elmo666's efforts spread, he began getting requests from drug dealers who wanted him to use their wares for testing. Clearly, they figured that a positive result from Energy Control would be a fantastic marketing tool to draw more customers. They even offered elmo666 free samples. (He passed.)

-

Meanwhile, some in the purchasing community are arguing that those running markets on the deep web should be providing quality control themselves. PlutoPete told me over the phone that he had been in discussions about this with Dread Pirate Roberts, the pseudonymous owner of the original Silk Road site. “We [had been] talking about that on a more organized basis on Silk Road 1, doing lots of anonymous buys to police each category. But of course they took the thing [Silk Road] down before we got it properly off the ground,” he lamented.

-

But perhaps it is best that the users, those who are actually consuming the drugs, remain in charge of shaming dealers and warning each other. “It’s our responsibility to police the market based on reviews and feedback,” elmo666 wrote in an Evolution forum post. It seems that in the lawless space of the deep web, where everything from child porn to weapons are sold openly, users have cooperated in an organic display of self-regulation to stamp out those particular batches of drugs that are more likely to harm users.

-

“That’s always been the case with the deep web,” PlutoPete told me. Indeed, ever since Silk Road, a stable of the drug markets has been the review system, where buyers can leave a rating and feedback for vendors, letting others know about the reliability of the seller. But DoctorX’s lab, rigorously testing the products with scientific instruments, takes it a step further.

-
-
-
-
-
Photo by Joan Bardeletti
-
-
-
-

“In the white market, they have quality control. In the dark market, it should be the same,” Cristina Gil Lladanosa says to me before I leave the Barcelona lab.

-

A week after I visit the lab, the results of the MDMA arrive in my inbox: it is 85 percent pure, with no indications of other active ingredients. Whoever ordered that sample from the digital shelves of the deep web, and had it shipped to their doorstep in Canada, got hold of some seriously good, and relatively safe drugs. And now they know it.

-
-
-
-

Top photo by Joan Bardeletti

-

Follow Backchannel: Twitter |Facebook

-
-
-
+
+
+
+
+
+
+
+

Welcome to DoctorX’s Barcelona lab, where the drugs you bought online are tested for safety and purity. No questions asked.

+
+
+
+

Standing at a table in a chemistry lab in Barcelona, Cristina Gil Lladanosa tears open a silver, smell-proof protective envelope. She slides out a transparent bag full of crystals. Around her, machines whir and hum, and other researchers mill around in long, white coats.

+

She is holding the lab’s latest delivery of a drug bought from the “deep web,” the clandestine corner of the internet that isn’t reachable by normal search engines, and is home to some sites that require special software to access. Labeled as MDMA (the street term is ecstasy), this sample has been shipped from Canada. Lladanosa and her colleague Iván Fornís Espinosa have also received drugs, anonymously, from people in China, Australia, Europe and the United States.

+

“Here we have speed, MDMA, cocaine, pills,” Lladanosa says, pointing to vials full of red, green, blue and clear solutions sitting in labeled boxes.

+
+
+
+
+
Cristina Gil Lladanosa, at the Barcelona testing lab | photo by Joan Bardeletti
+
+
+
+

Since 2011, with the launch of Silk Road, anybody has been able to safely buy illegal drugs from the deep web and have them delivered to their door. Though the FBI shut down that black market in October 2013, other outlets have emerged to fill its role. For the last 10 months the lab at which Lladanosa and Espinosa work has offered a paid testing service of those drugs. By sending in samples for analysis, users can know exactly what it is they are buying, and make a more informed decision about whether to ingest the substance. The group, called Energy Control, which has being running “harm reduction” programs since 1999, is the first to run a testing service explicitly geared towards verifying those purchases from the deep web.

+

Before joining Energy Control, Lladanosa briefly worked at a pharmacy, whereas Espinosa spent 14 years doing drug analysis. Working at Energy Control is “more gratifying,” and “rewarding” than her previous jobs, Lladanosa told me. They also receive help from a group of volunteers, made up of a mixture of “squatters,” as Espinosa put it, and medical students, who prepare the samples for testing.

+

After weighing out the crystals, aggressively mixing it with methanol until dissolved, and delicately pouring the liquid into a tiny brown bottle, Lladanosa, a petite woman who is nearly engulfed by her lab coat, is now ready to test the sample. She loads a series of three trays on top of a large white appliance sitting on a table, called a gas chromatograph (GC). A jungle of thick pipes hang from the lab’s ceiling behind it.

+
+
+
+
+
Photo by Joan Bardeletti
+
+
+
+

“Chromatography separates all the substances,” Lladanosa says as she loads the machine with an array of drugs sent from the deep web and local Spanish users. It can tell whether a sample is pure or contaminated, and if the latter, with what.

+

Rushes of hot air blow across the desk as the gas chromatograph blasts the sample at 280 degrees Celsius. Thirty minutes later the machine’s robotic arm automatically moves over to grip another bottle. The machine will continue cranking through the 150 samples in the trays for most of the work week.

+
+
+
+
+
Photo by Joan Bardeletti
+
+
+
+

To get the drugs to Barcelona, a user mails at least 10 milligrams of a substance to the offices of the Asociación Bienestar y Desarrollo, the non-government organization that oversees Energy Control. The sample then gets delivered to the testing service’s laboratory, at the Barcelona Biomedical Research Park, a futuristic, seven story building sitting metres away from the beach. Energy Control borrows its lab space from a biomedical research group for free.

+

The tests cost 50 Euro per sample. Users pay, not surprisingly, with Bitcoin. In the post announcing Energy Control’s service on the deep web, the group promised that “All profits of this service are set aside of maintenance of this project.”

+

About a week after testing, those results are sent in a PDF to an email address provided by the anonymous client.

+

“The process is quite boring, because you are in a routine,” Lladanosa says. But one part of the process is consistently surprising: that moment when the results pop up on the screen. “Every time it’s something different.” For instance, one cocaine sample she had tested also contained phenacetin, a painkiller added to increase the product’s weight; lidocaine, an anesthetic that numbs the gums, giving the impression that the user is taking higher quality cocaine; and common caffeine.

+
+
+
+

The deep web drug lab is the brainchild of Fernando Caudevilla, a Spanish physician who is better known as “DoctorX” on the deep web, a nickname given to him by his Energy Control co-workers because of his earlier writing about the history, risks and recreational culture of MDMA. In the physical world, Caudevilla has worked for over a decade with Energy Control on various harm reduction focused projects, most of which have involved giving Spanish illegal drug users medical guidance, and often writing leaflets about the harms of certain substances.

+
+
+
+
+
Fernando Caudevilla, AKA DoctorX. Photo: Joseph Cox
+
+
+
+

Caudevilla first ventured into Silk Road forums in April 2013. “I would like to contribute to this forum offering professional advice in topics related to drug use and health,” he wrote in an introductory post, using his DoctorX alias. Caudevilla offered to provide answers to questions that a typical doctor is not prepared, or willing, to respond to, at least not without a lecture or a judgment. “This advice cannot replace a complete face-to-face medical evaluation,” he wrote, “but I know how difficult it can be to talk frankly about these things.”

+

The requests flooded in. A diabetic asked what effect MDMA has on blood sugar; another what the risks of frequent psychedelic use were for a young person. Someone wanted to know whether amphetamine use should be avoided during lactation. In all, Fernando’s thread received over 50,000 visits and 300 questions before the FBI shut down Silk Road.

+

“He’s amazing. A gift to this community,” one user wrote on the Silk Road 2.0 forum, a site that sprang up after the original. “His knowledge is invaluable, and never comes with any judgment.” Up until recently, Caudevilla answered questions on the marketplace “Evolution.” Last week, however, the administrators of that site pulled a scam, shutting the market down and escaping with an estimated $12 million worth of Bitcoin.

+

Caudevilla’s transition from dispensing advice to starting up a no-questions-asked drug testing service came as a consequence of his experience on the deep web. He’d wondered whether he could help bring more harm reduction services to a marketplace without controls. The Energy Control project, as part of its mandate of educating drug users and preventing harm, had already been carrying out drug testing for local Spanish users since 2001, at music festivals, night clubs, or through a drop-in service at a lab in Madrid.

+

“I thought, we are doing this in Spain, why don’t we do an international drug testing service?” Caudevilla told me when I visited the other Energy Control lab, in Madrid. Caudevilla, a stocky character with ear piercings and short, shaved hair, has eyes that light up whenever he discusses the world of the deep web. Later, via email, he elaborated that it was not a hard sell. “It was not too hard to convince them,” he wrote me. Clearly, Energy Control believed that the reputation he had earned as an unbiased medical professional on the deep web might carry over to the drug analysis service, where one needs to establish “credibility, trustworthiness, [and] transparency,” Caudevilla said. “We could not make mistakes,” he added.

+
+
+
+
+
Photo: Joseph Cox
+
+
+
+
+
+
+

While the Energy Control lab in Madrid lab only tests Spanish drugs from various sources, it is the Barcelona location which vets the substances bought in the shadowy recesses of of the deep web. Caudevilla no longer runs it, having handed it over to his colleague Ana Muñoz. She maintains a presence on the deep web forums, answers questions from potential users, and sends back reports when they are ready.

+

The testing program exists in a legal grey area. The people who own the Barcelona lab are accredited to experiment with and handle drugs, but Energy Control doesn’t have this permission itself, at least not in writing.

+

“We have a verbal agreement with the police and other authorities. They already know what we are doing,” Lladanosa tells me. It is a pact of mutual benefit. Energy Control provides the police with information on batches of drugs in Spain, whether they’re from the deep web or not, Espinosa says. They also contribute to the European Monitoring Centre for Drugs and Drug Addiction’s early warning system, a collaboration that attempts to spread information about dangerous drugs as quickly as possible.

+

By the time of my visit in February, Energy Control had received over 150 samples from the deep web and have been receiving more at a rate of between 4 and 8 a week. Traditional drugs, such as cocaine and MDMA, make up about 70 percent of the samples tested, but the Barcelona lab has also received samples of the prescription pill codeine, research chemicals and synthetic cannabinoids, and even pills of Viagra.

+
+
+
+
+
Photo by Joan Bardeletti
+
+
+
+

So it’s fair to make a tentative judgement on what people are paying for on the deep web. The verdict thus far? Overall, drugs on the deep web appear to be of much higher quality than those found on the street.

+

“In general, the cocaine is amazing,” says Caudevilla, saying that the samples they’ve seen have purities climbing towards 80 or 90 percent, and some even higher. To get an idea of how unusual this is, take a look at the United Nations Office on Drugs and Crime World Drug Report 2014, which reports that the average quality of street cocaine in Spain is just over 40 percent, while in the United Kingdom it is closer to 30 percent.“We have found 100 percent [pure] cocaine,” he adds. “That’s really, really strange. That means that, technically, this cocaine has been purified, with clandestine methods.”

+

Naturally, identifying vendors who sell this top-of-the-range stuff is one of the reasons that people have sent samples to Energy Control. Caudevilla was keen to stress that, officially, Energy Control’s service “is not intended to be a control of drug quality,” meaning a vetting process for identifying the best sellers, but that is exactly how some people have been using it.

+

As one buyer on the Evolution market, elmo666, wrote to me over the site’s messaging system, “My initial motivations were selfish. My primary motivation was to ensure that I was receiving and continue to receive a high quality product, essentially to keep the vendor honest as far as my interactions with them went.”

+

Vendors on deep web markets advertise their product just like any other outlet does, using flash sales, gimmicky giveaways and promises of drugs that are superior to those of their competitors. The claims, however, can turn out to be empty: despite the test results that show that deep web cocaine vendors typically sell product that is of a better quality than that found on the street, in plenty of cases, the drugs are nowhere near as pure as advertised.

+

“You won’t be getting anything CLOSE to what you paid for,” one user complained about the cocaine from ‘Mirkov’, a vendor on Evolution. “He sells 65% not 95%.”

+
+
+
+
+
Photo by Joan Bardeletti
+
+
+
+
+
+
+

Despite the prevalence of people using the service to gauge the quality of what goes up their nose, many users send samples to Energy Control in the spirit of its original mission: keeping themselves alive and healthy. The worst case scenario from drugs purchased on the deep web is, well the worst case. That was the outcome when Patrick McMullen, a 17-year-old Scottish student, ingested half a gram of MDMA and three tabs of LSD, reportedly purchased from the Silk Road. While talking to his friends on Skype, his words became slurred and he passed out. Paramedics could not revive him. The coroner for that case, Sherrif Payne, who deemed the cause of death ecstasy toxicity, told The Independent “You never know the purity of what you are taking and you can easily come unstuck.”

+

ScreamMyName, a deep web user who has been active since the original Silk Road, wants to alert users to the dangerous chemicals that are often mixed with drugs, and is using Energy Control as a means to do so.

+

“We’re at a time where some vendors are outright sending people poison. Some do it unknowingly,” ScreamMyName told me in an encrypted message. “Cocaine production in South America is often tainted with either levamisole or phenacetine. Both poison to humans and both with severe side effects.”

+

In the case of Levamisole, those prescribing it are often not doctors but veterinarians, as Levamisole is commonly used on animals, primarily for the treatment of worms. If ingested by humans it can lead to cases of extreme eruptions of the skin, as documented in a study from researchers at the University of California, San Francisco. But Lladanosa has found Levamisole in cocaine samples; dealers use it to increase the product weight, allowing them to stretch their batch further for greater profit — and also, she says, because Levamisole has a strong stimulant effect.

+

“It got me sick as fuck,” Dr. Feel, an Evolution user, wrote on the site’s forums after consuming cocaine that had been cut with 23 percent Levamisole, and later tested by Energy Control. “I was laid up in bed for several days because of that shit. The first night I did it, I thought I was going to die. I nearly drove myself to the ER.”

+

“More people die because of tainted drugs than the drugs themselves,” Dr. Feel added. “It’s the cuts and adulterants that are making people sick and killing them.”

+
+
+
+
+
Photo by Joan Bardeletti
+
+
+
+

The particular case of cocaine cut with Levamisole is one of the reasons that ScreamMyName has been pushing for more drug testing on the deep web markets. “I recognize that drug use isn’t exactly healthy, but why exacerbate the problem?” he told me when I contacted him after his post. “[Energy Control] provides a way for users to test the drugs they’ll use and for these very users to know what it is they’re putting in their bodies. Such services are in very short supply.”

+

After sending a number of Energy Control tests himself, ScreamMyName started a de facto crowd-sourcing campaign to get more drugs sent to the lab, and then shared the results, after throwing in some cash to get the ball rolling. He set up a Bitcoin wallet, with the hope that users might chip in to fund further tests. At the time of writing, the wallet has received a total of 1.81 bitcoins; around $430 at today’s exchange rates.

+

In posts to the Evolution community, ScreamMyName pitched this project as something that will benefit users and keep drug dealer honest. “When the funds build up to a point where we can purchase an [Energy Control] test fee, we’ll do a US thread poll for a few days and try to cohesively decide on what vendor to test,” he continued.

+
+
+
+
+
Photo by Joan Bardeletti
+
+
+
+

Other members of the community have been helping out, too. PlutoPete, a vendor from the original Silk Road who sold cannabis seeds and other legal items, has provided ScreamMyName with packaging to safely send the samples to Barcelona. “A box of baggies, and a load of different moisture barrier bags,” PlutoPete told me over the phone. “That’s what all the vendors use.”

+

It’s a modest program so far. ScreamMyName told me that so far he had gotten enough public funding to purchase five different Energy Control tests, in addition to the ten or so he’s sent himself so far. “The program created is still in its infancy and it is growing and changing as we go along but I have a lot of faith in what we’re doing,” he says.

+

But the spirit is contagious: elmo666, the other deep web user testing cocaine, originally kept the results of the drug tests to himself, but he, too, saw a benefit to distributing the data. “It is clear that it is a useful service to other users, keeping vendors honest and drugs (and their users) safe,” he told me. He started to report his findings to others on the forums, and then created a thread with summaries of the test results, as well as comments from the vendors if they provided it. Other users were soon basing their decisions on what to buy on elmo666‘s tests.

+

“I’m defo trying the cola based on the incredibly helpful elmo and his energy control results and recommendations,” wrote user jayk1984. On top of this, elmo666 plans to launch an independent site on the deep web that will collate all of these results, which should act as a resource for users of all the marketplaces.

+

As word of elmo666's efforts spread, he began getting requests from drug dealers who wanted him to use their wares for testing. Clearly, they figured that a positive result from Energy Control would be a fantastic marketing tool to draw more customers. They even offered elmo666 free samples. (He passed.)

+

Meanwhile, some in the purchasing community are arguing that those running markets on the deep web should be providing quality control themselves. PlutoPete told me over the phone that he had been in discussions about this with Dread Pirate Roberts, the pseudonymous owner of the original Silk Road site. “We [had been] talking about that on a more organized basis on Silk Road 1, doing lots of anonymous buys to police each category. But of course they took the thing [Silk Road] down before we got it properly off the ground,” he lamented.

+

But perhaps it is best that the users, those who are actually consuming the drugs, remain in charge of shaming dealers and warning each other. “It’s our responsibility to police the market based on reviews and feedback,” elmo666 wrote in an Evolution forum post. It seems that in the lawless space of the deep web, where everything from child porn to weapons are sold openly, users have cooperated in an organic display of self-regulation to stamp out those particular batches of drugs that are more likely to harm users.

+

“That’s always been the case with the deep web,” PlutoPete told me. Indeed, ever since Silk Road, a stable of the drug markets has been the review system, where buyers can leave a rating and feedback for vendors, letting others know about the reliability of the seller. But DoctorX’s lab, rigorously testing the products with scientific instruments, takes it a step further.

+
+
+
+
+
Photo by Joan Bardeletti
+
+
+
+

“In the white market, they have quality control. In the dark market, it should be the same,” Cristina Gil Lladanosa says to me before I leave the Barcelona lab.

+

A week after I visit the lab, the results of the MDMA arrive in my inbox: it is 85 percent pure, with no indications of other active ingredients. Whoever ordered that sample from the digital shelves of the deep web, and had it shipped to their doorstep in Canada, got hold of some seriously good, and relatively safe drugs. And now they know it.

+
+
+
+

Top photo by Joan Bardeletti

+

Follow Backchannel: Twitter |Facebook

+
\ No newline at end of file diff --git a/test/test-pages/lwn-1/expected.html b/test/test-pages/lwn-1/expected.html index 9c81978..8dc604a 100644 --- a/test/test-pages/lwn-1/expected.html +++ b/test/test-pages/lwn-1/expected.html @@ -1,581 +1,576 @@
-
- -
-

A trademark battle in the Arduino community

-

The Arduino has been one of the biggest success stories of the open-hardware movement, but that success does not protect it from internal conflict. In recent months, two of the project's founders have come into conflict about the direction of future efforts—and that conflict has turned into a legal dispute about who owns the rights to the Arduino trademark.

-

The current fight is a battle between two companies that both bear the Arduino name: Arduino LLC and Arduino SRL. The disagreements that led to present state of affairs go back a bit further.

-

The Arduino project grew out of 2005-era course work taught at the Interaction Design Institute Ivrea (IDII) in Ivrea, Italy (using Processing, Wiring, and pre-existing microcontroller hardware). After the IDII program was discontinued, the open-hardware Arduino project as we know it was launched by Massimo Banzi, David Cuartielles, and David Mellis (who had worked together at IDII), with co-founders Tom Igoe and Gianluca Martino joining shortly afterward. The project released open hardware designs (including full schematics and design files) as well as the microcontroller software to run on the boards and the desktop IDE needed to program it.

-

Arduino LLC was incorporated in 2008 by Banzi, Cuartielles, Mellis, Igoe, and Martino. The company is registered in the United States, and it has continued to design the Arduino product line, develop the software, and run the Arduino community site. The hardware devices themselves, however, were manufactured by a separate company, "Smart Projects SRL," that was founded by Martino. "SRL" is essentially the Italian equivalent of "LLC"—Smart Projects was incorporated in Italy.

-

This division of responsibilities—with the main Arduino project handling everything except for board manufacturing—may seem like an odd one, but it is consistent with Arduino's marketing story. From its earliest days, the designs for the hardware have been freely available, and outside companies were allowed to make Arduino-compatible devices. The project has long run a certification +

+

A trademark battle in the Arduino community

+

The Arduino has been one of the biggest success stories of the open-hardware movement, but that success does not protect it from internal conflict. In recent months, two of the project's founders have come into conflict about the direction of future efforts—and that conflict has turned into a legal dispute about who owns the rights to the Arduino trademark.

+

The current fight is a battle between two companies that both bear the Arduino name: Arduino LLC and Arduino SRL. The disagreements that led to present state of affairs go back a bit further.

+

The Arduino project grew out of 2005-era course work taught at the Interaction Design Institute Ivrea (IDII) in Ivrea, Italy (using Processing, Wiring, and pre-existing microcontroller hardware). After the IDII program was discontinued, the open-hardware Arduino project as we know it was launched by Massimo Banzi, David Cuartielles, and David Mellis (who had worked together at IDII), with co-founders Tom Igoe and Gianluca Martino joining shortly afterward. The project released open hardware designs (including full schematics and design files) as well as the microcontroller software to run on the boards and the desktop IDE needed to program it.

+

Arduino LLC was incorporated in 2008 by Banzi, Cuartielles, Mellis, Igoe, and Martino. The company is registered in the United States, and it has continued to design the Arduino product line, develop the software, and run the Arduino community site. The hardware devices themselves, however, were manufactured by a separate company, "Smart Projects SRL," that was founded by Martino. "SRL" is essentially the Italian equivalent of "LLC"—Smart Projects was incorporated in Italy.

+

This division of responsibilities—with the main Arduino project handling everything except for board manufacturing—may seem like an odd one, but it is consistent with Arduino's marketing story. From its earliest days, the designs for the hardware have been freely available, and outside companies were allowed to make Arduino-compatible devices. The project has long run a certification program for third-party manufacturers interested in using the "Arduino" branding, but allows (and arguably even encourages) informal software and firmware compatibility.

-

The Arduino branding was not formally registered as a trademark in the early days, however. Arduino LLC filed to register the US trademark in April 2009, and it was granted in 2011.

-

At this point, the exact events begin to be harder to verify, but the original group of founders reportedly had a difference of opinion about how to license out hardware production rights to other companies. Wired Italy reports that Martino and Smart Projects resisted the other four founders' plans to "internationalize" production—although it is not clear if that meant that Smart Projects disapproved of licensing out any official hardware manufacturing to other companies, or had some other concern. Heise Online adds that the conflict seemed to be about moving some production to China.

-

What is clear is that Smart Projects filed a petition with the US Patent and Trademark Office (USPTO) in October 2014 asking the USPTO to cancel Arduino LLC's trademark on "Arduino." Then, in November 2014, Smart Projects changed its company's name to Arduino SRL. Somewhere around that time, Martino sold off his ownership stake in Smart Projects SRL and new owner Federico Musto was named CEO.

-

Unsurprisingly, Arduino LLC did not care for the petition to the USPTO and, in January 2015, the company filed a trademark-infringement lawsuit against Arduino SRL. Confusing matters further, the re-branded Arduino SRL has set up its own web site using the domain name arduino.org, which duplicates most of the site features found on the original Arduino site (arduino.cc). That includes both a hardware store and software downloads.

-

Musto, the new CEO of the company now called Arduino SRL, has a bit of a history with Arduino as well. His other manufacturing business had collaborated with Arduino LLC on the design and production of the Arduino Yún, which has received some criticism for including proprietary components.

-

Hackaday has run a two-part series (in February and March) digging into the ins and outs of the dispute, including the suggestion that Arduino LLC's recent release of version 1.6.0 of the Arduino IDE was a move intended to block Arduino SRL from hijacking IDE development. Commenter Paul Stoffregen (who was the author of the Heise story above) noted that Arduino SRL recently created a fork of the Arduino IDE on GitHub.

-

Most recently, Banzi broke his silence about the dispute in a story published at MAKEzine. There, Banzi claims that Martino secretly filed a trademark application on "Arduino" in Italy in 2008 and told none of the other Arduino founders. He also details a series of unpleasant negotiations between the companies, including Smart Projects stopping the royalty payments it had long sent to Arduino LLC for manufacturing devices and re-branding its boards with the Arduino.org URL.

-

Users appear to be stuck in the middle. Banzi says that several retail outlets that claim to be selling "official" Arduino boards are actually paying Arduino SRL, not Arduino LLC, but it is quite difficult to determine which retailers are lined up on which side, since there are (typically) several levels of supplier involved. The two Arduino companies' web sites also disagree about the available hardware, with Arduino.org offering the new Arduino Zero model for sale today and Arduino.cc listing it as "Coming soon."

-

Furthermore, as Hackaday's March story explains, the recently-released Arduino.cc IDE now reports that boards manufactured by Arduino SRL are "uncertified." That warning does not prevent users from programming the other company's hardware, but it will no doubt confuse quite a few users who believe they possess genuine Arduino-manufactured devices.

-

The USPTO page for Arduino SRL's petition notes pre-trial disclosure dates have been set for August and October of 2015 (for Arduino SRL and Arduino LLC, respectively), which suggests that this debate is far from over. Of course, it is always disappointing to observe a falling out between project founders, particularly when the project in question has had such an impact on open-source software and open hardware.

-

One could argue that disputes of this sort are proof that even small projects started among friends need to take legal and intellectual-property issues (such as trademarks) seriously from the very beginning—perhaps Arduino and Smart Projects thought that an informal agreement was all that was necessary in the early days, after all.

-

But, perhaps, once a project becomes profitable, there is simply no way to predict what might happen. Arduino LLC would seem to have a strong case for continual and rigorous use of the "Arduino" trademark, which is the salient point in US trademark law. It could still be a while before the courts rule on either side of that question, however.

-

Comments (5 posted)

-

Mapping and data mining with QGIS 2.8

-

By Nathan Willis -
March 25, 2015

-

QGIS is a free-software geographic information system (GIS) tool; it provides a unified interface in which users can import, edit, and analyze geographic-oriented information, and it can produce output as varied as printable maps or map-based web services. The project recently made its first update to be designated a long-term release (LTR), and that release is both poised for high-end usage and friendly to newcomers alike.

-

The new release is version 2.8, which was unveiled on March 2. An official change +

The Arduino branding was not formally registered as a trademark in the early days, however. Arduino LLC filed to register the US trademark in April 2009, and it was granted in 2011.

+

At this point, the exact events begin to be harder to verify, but the original group of founders reportedly had a difference of opinion about how to license out hardware production rights to other companies. Wired Italy reports that Martino and Smart Projects resisted the other four founders' plans to "internationalize" production—although it is not clear if that meant that Smart Projects disapproved of licensing out any official hardware manufacturing to other companies, or had some other concern. Heise Online adds that the conflict seemed to be about moving some production to China.

+

What is clear is that Smart Projects filed a petition with the US Patent and Trademark Office (USPTO) in October 2014 asking the USPTO to cancel Arduino LLC's trademark on "Arduino." Then, in November 2014, Smart Projects changed its company's name to Arduino SRL. Somewhere around that time, Martino sold off his ownership stake in Smart Projects SRL and new owner Federico Musto was named CEO.

+

Unsurprisingly, Arduino LLC did not care for the petition to the USPTO and, in January 2015, the company filed a trademark-infringement lawsuit against Arduino SRL. Confusing matters further, the re-branded Arduino SRL has set up its own web site using the domain name arduino.org, which duplicates most of the site features found on the original Arduino site (arduino.cc). That includes both a hardware store and software downloads.

+

Musto, the new CEO of the company now called Arduino SRL, has a bit of a history with Arduino as well. His other manufacturing business had collaborated with Arduino LLC on the design and production of the Arduino Yún, which has received some criticism for including proprietary components.

+

Hackaday has run a two-part series (in February and March) digging into the ins and outs of the dispute, including the suggestion that Arduino LLC's recent release of version 1.6.0 of the Arduino IDE was a move intended to block Arduino SRL from hijacking IDE development. Commenter Paul Stoffregen (who was the author of the Heise story above) noted that Arduino SRL recently created a fork of the Arduino IDE on GitHub.

+

Most recently, Banzi broke his silence about the dispute in a story published at MAKEzine. There, Banzi claims that Martino secretly filed a trademark application on "Arduino" in Italy in 2008 and told none of the other Arduino founders. He also details a series of unpleasant negotiations between the companies, including Smart Projects stopping the royalty payments it had long sent to Arduino LLC for manufacturing devices and re-branding its boards with the Arduino.org URL.

+

Users appear to be stuck in the middle. Banzi says that several retail outlets that claim to be selling "official" Arduino boards are actually paying Arduino SRL, not Arduino LLC, but it is quite difficult to determine which retailers are lined up on which side, since there are (typically) several levels of supplier involved. The two Arduino companies' web sites also disagree about the available hardware, with Arduino.org offering the new Arduino Zero model for sale today and Arduino.cc listing it as "Coming soon."

+

Furthermore, as Hackaday's March story explains, the recently-released Arduino.cc IDE now reports that boards manufactured by Arduino SRL are "uncertified." That warning does not prevent users from programming the other company's hardware, but it will no doubt confuse quite a few users who believe they possess genuine Arduino-manufactured devices.

+

The USPTO page for Arduino SRL's petition notes pre-trial disclosure dates have been set for August and October of 2015 (for Arduino SRL and Arduino LLC, respectively), which suggests that this debate is far from over. Of course, it is always disappointing to observe a falling out between project founders, particularly when the project in question has had such an impact on open-source software and open hardware.

+

One could argue that disputes of this sort are proof that even small projects started among friends need to take legal and intellectual-property issues (such as trademarks) seriously from the very beginning—perhaps Arduino and Smart Projects thought that an informal agreement was all that was necessary in the early days, after all.

+

But, perhaps, once a project becomes profitable, there is simply no way to predict what might happen. Arduino LLC would seem to have a strong case for continual and rigorous use of the "Arduino" trademark, which is the salient point in US trademark law. It could still be a while before the courts rule on either side of that question, however.

+

Comments (5 posted)

+

Mapping and data mining with QGIS 2.8

+

By Nathan Willis +
March 25, 2015

+

QGIS is a free-software geographic information system (GIS) tool; it provides a unified interface in which users can import, edit, and analyze geographic-oriented information, and it can produce output as varied as printable maps or map-based web services. The project recently made its first update to be designated a long-term release (LTR), and that release is both poised for high-end usage and friendly to newcomers alike.

+

The new release is version 2.8, which was unveiled on March 2. An official change log is available on the QGIS site, while the release itself was announced primarily through blog posts (such as this post by Anita Graser of the project's steering committee). Downloads are available for a variety of platforms, including packages for Ubuntu, Debian, Fedora, openSUSE, and several other distributions.

- [QGIS main interface] -

As the name might suggest, QGIS is a Qt application; the latest release will, in fact, build on both Qt4 and Qt5, although the binaries released by the project come only in Qt4 form at present. 2.8 has been labeled a long-term release (LTR)—which, in this case, means that the project has committed to providing backported bug fixes for one full calendar year, and that the 2.8.x series is in permanent feature freeze. The goal, according to the change log, is to provide a stable version suitable for businesses and deployments in other large organizations. The change log itself points out that the development of quite a few new features was underwritten by various GIS companies or university groups, which suggests that taking care of these organizations' needs is reaping dividends for the project.

-

For those new to QGIS (or GIS in general), there is a detailed new-user tutorial that provides a thorough walk-through of the data-manipulation, mapping, and analysis functions. Being a new user, I went through the tutorial; although there are a handful of minor differences between QGIS 2.8 and the version used in the text (primarily whether specific features were accessed through a toolbar or right-click menu), on the whole it is well worth the time.

-

QGIS is designed to make short work of importing spatially oriented data sets, mining information from them, and turning the results into a meaningful visualization. Technically speaking, the visualization output is optional: one could simply extract the needed statistics and results and use them to answer some question or, perhaps, publish the massaged data set as a database for others to use.

-

But well-made maps are often the easiest way to illuminate facts about populations, political regions, geography, and many other topics when human comprehension is the goal. QGIS makes importing data from databases, web-mapping services (WMS), and even unwieldy flat-file data dumps a painless experience. It handles converting between a variety of map-referencing systems more or less automatically, and allows the user to focus on finding the useful attributes of the data sets and rendering them on screen.

-

Here be data

-

The significant changes in QGIS 2.8 fall into several categories. There are updates to how QGIS handles the mathematical expressions and queries users can use to filter information out of a data set, improvements to the tools used to explore the on-screen map canvas, and enhancements to the "map composer" used to produce visual output. This is on top of plenty of other under-the-hood improvements, naturally.

- [QGIS query builder] -

In the first category are several updates to the filtering tools used to mine a data set. Generally speaking, each independent data set is added to a QGIS project as its own layer, then transformed with filters to focus in on a specific portion of the original data. For instance, the land-usage statistics for a region might be one layer, while roads and buildings for the same region from OpenStreetMap might be two additional layers. Such filters can be created in several ways: there is a "query builder" that lets the user construct and test expressions on a data layer, then save the results, an SQL console for performing similar queries on a database, and spreadsheet-like editing tools for working directly on data tables.

-

All three have been improved in this release. New are support for if(condition, true, false) conditional statements, a set of operations for geometry primitives (e.g., to test whether regions overlap or lines intersect), and an "integer divide" operation. Users can also add comments to their queries to annotate their code, and there is a new custom + [QGIS main interface] +

As the name might suggest, QGIS is a Qt application; the latest release will, in fact, build on both Qt4 and Qt5, although the binaries released by the project come only in Qt4 form at present. 2.8 has been labeled a long-term release (LTR)—which, in this case, means that the project has committed to providing backported bug fixes for one full calendar year, and that the 2.8.x series is in permanent feature freeze. The goal, according to the change log, is to provide a stable version suitable for businesses and deployments in other large organizations. The change log itself points out that the development of quite a few new features was underwritten by various GIS companies or university groups, which suggests that taking care of these organizations' needs is reaping dividends for the project.

+

For those new to QGIS (or GIS in general), there is a detailed new-user tutorial that provides a thorough walk-through of the data-manipulation, mapping, and analysis functions. Being a new user, I went through the tutorial; although there are a handful of minor differences between QGIS 2.8 and the version used in the text (primarily whether specific features were accessed through a toolbar or right-click menu), on the whole it is well worth the time.

+

QGIS is designed to make short work of importing spatially oriented data sets, mining information from them, and turning the results into a meaningful visualization. Technically speaking, the visualization output is optional: one could simply extract the needed statistics and results and use them to answer some question or, perhaps, publish the massaged data set as a database for others to use.

+

But well-made maps are often the easiest way to illuminate facts about populations, political regions, geography, and many other topics when human comprehension is the goal. QGIS makes importing data from databases, web-mapping services (WMS), and even unwieldy flat-file data dumps a painless experience. It handles converting between a variety of map-referencing systems more or less automatically, and allows the user to focus on finding the useful attributes of the data sets and rendering them on screen.

+

Here be data

+

The significant changes in QGIS 2.8 fall into several categories. There are updates to how QGIS handles the mathematical expressions and queries users can use to filter information out of a data set, improvements to the tools used to explore the on-screen map canvas, and enhancements to the "map composer" used to produce visual output. This is on top of plenty of other under-the-hood improvements, naturally.

+ [QGIS query builder] +

In the first category are several updates to the filtering tools used to mine a data set. Generally speaking, each independent data set is added to a QGIS project as its own layer, then transformed with filters to focus in on a specific portion of the original data. For instance, the land-usage statistics for a region might be one layer, while roads and buildings for the same region from OpenStreetMap might be two additional layers. Such filters can be created in several ways: there is a "query builder" that lets the user construct and test expressions on a data layer, then save the results, an SQL console for performing similar queries on a database, and spreadsheet-like editing tools for working directly on data tables.

+

All three have been improved in this release. New are support for if(condition, true, false) conditional statements, a set of operations for geometry primitives (e.g., to test whether regions overlap or lines intersect), and an "integer divide" operation. Users can also add comments to their queries to annotate their code, and there is a new custom function editor for writing Python functions that can be called in mathematical expressions within the query builder.

-

It is also now possible to select only some rows in a table, then perform calculations just on the selection—previously, users would have to extract the rows of interest into a new table first. Similarly, in the SQL editor, the user can highlight a subset of the SQL query and execute it separately, which is no doubt helpful for debugging.

-

There have also been several improvements to the Python and Processing plugins. Users can now drag-and-drop Python scripts onto QGIS and they will be run automatically. Several new analysis algorithms are now available through the Processing interface that were previously Python-only; they include algorithms for generating grids of points or vectors within a region, splitting layers and lines, generating hypsometric +

It is also now possible to select only some rows in a table, then perform calculations just on the selection—previously, users would have to extract the rows of interest into a new table first. Similarly, in the SQL editor, the user can highlight a subset of the SQL query and execute it separately, which is no doubt helpful for debugging.

+

There have also been several improvements to the Python and Processing plugins. Users can now drag-and-drop Python scripts onto QGIS and they will be run automatically. Several new analysis algorithms are now available through the Processing interface that were previously Python-only; they include algorithms for generating grids of points or vectors within a region, splitting layers and lines, generating hypsometric curves, refactoring data sets, and more.

-

Maps in, maps out

- [QGIS simplify tool] -

The process of working with on-screen map data picked up some improvements in the new release as well. Perhaps the most fundamental is that each map layer added to the canvas is now handled in its own thread, so fewer hangs in the user interface are experienced when re-rendering a layer (as happens whenever the user changes the look of points or shapes in a layer). Since remote databases can also be layers, this multi-threaded approach is more resilient against connectivity problems, too. The interface also now supports temporary "scratch" layers that can be used to merge, filter, or simply experiment with a data set, but are not saved when the current project is saved.

-

For working on the canvas itself, polygonal regions can now use raster images (tiled, if necessary) as fill colors, the map itself can be rotated arbitrarily, and objects can be "snapped" to align with items on any layer (not just the current layer). For working with raster image layers (e.g., aerial photographs) or simply creating new geometric shapes by hand, there is a new digitizing tool that can offer assistance by locking lines to specific angles, automatically keeping borders parallel, and other niceties.

-

There is a completely overhauled "simplify" tool that is used to reduce the number of extraneous vertices of a vector layer (thus reducing its size). The old simplify tool provided only a relative "tolerance" setting that did not correspond directly to any units. With the new tool, users can set a simplification threshold in terms of the underlying map units, layer-specific units, pixels, and more—and, in addition, the tool reports how much the simplify operation has reduced the size of the data.

- [QGIS style editing] -

There has also been an effort to present a uniform interface to one of the most important features of the map canvas: the ability to change the symbology used for an item based on some data attribute. The simplest example might be to change the line color of a road based on whether its road-type attribute is "highway," "service road," "residential," or so on. But the same feature is used to automatically highlight layer information based on the filtering and querying functionality discussed above. The new release allows many more map attributes to be controlled by these "data definition" settings, and provides a hard-to-miss button next to each attribute, through which a custom data definition can be set.

-

QGIS's composer module is the tool used to take project data and generate a map that can be used outside of the application (in print, as a static image, or as a layer for MapServer or some other software tool, for example). Consequently, it is not a simple select-and-click-export tool; composing the output can involve a lot of choices about which data to make visible, how (and where) to label it, and how to make it generally accessible.

-

The updated composer in 2.8 now has a full-screen mode and sports several new options for configuring output. For instance, the user now has full control over how map axes are labeled. In previous releases, the grid coordinates of the map could be turned on or off, but the only options were all or nothing. Now, the user can individually choose whether coordinates are displayed on all four sides, and can even choose in which direction vertical text labels will run (so that they can be correctly justified to the edge of the map, for example).

-

There are, as usual, many more changes than there is room to discuss. Some particularly noteworthy improvements include the ability to save and load bookmarks for frequently used data sources (perhaps most useful for databases, web services, and other non-local data) and improvements to QGIS's server module. This module allows one QGIS instance to serve up data accessible to other QGIS applications (for example, to simply team projects). The server can now be extended with Python plugins and the data layers that it serves can be styled with style rules like those used in the desktop interface.

-

QGIS is one of those rare free-software applications that is both powerful enough for high-end work and yet also straightforward to use for the simple tasks that might attract a newcomer to GIS in the first place. The 2.8 release, particularly with its project-wide commitment to long-term support, appears to be an update well worth checking out, whether one needs to create a simple, custom map or to mine a database for obscure geo-referenced meaning.

-

Comments (3 posted)

-

Development activity in LibreOffice and OpenOffice

-

By Jonathan Corbet -
March 25, 2015

-

The LibreOffice project was

announced -

with great fanfare in September 2010. Nearly one year later, the OpenOffice.org project (from which LibreOffice was forked)

was +

Maps in, maps out

+
[QGIS simplify tool] +

The process of working with on-screen map data picked up some improvements in the new release as well. Perhaps the most fundamental is that each map layer added to the canvas is now handled in its own thread, so fewer hangs in the user interface are experienced when re-rendering a layer (as happens whenever the user changes the look of points or shapes in a layer). Since remote databases can also be layers, this multi-threaded approach is more resilient against connectivity problems, too. The interface also now supports temporary "scratch" layers that can be used to merge, filter, or simply experiment with a data set, but are not saved when the current project is saved.

+

For working on the canvas itself, polygonal regions can now use raster images (tiled, if necessary) as fill colors, the map itself can be rotated arbitrarily, and objects can be "snapped" to align with items on any layer (not just the current layer). For working with raster image layers (e.g., aerial photographs) or simply creating new geometric shapes by hand, there is a new digitizing tool that can offer assistance by locking lines to specific angles, automatically keeping borders parallel, and other niceties.

+

There is a completely overhauled "simplify" tool that is used to reduce the number of extraneous vertices of a vector layer (thus reducing its size). The old simplify tool provided only a relative "tolerance" setting that did not correspond directly to any units. With the new tool, users can set a simplification threshold in terms of the underlying map units, layer-specific units, pixels, and more—and, in addition, the tool reports how much the simplify operation has reduced the size of the data.

+ [QGIS style editing] +

There has also been an effort to present a uniform interface to one of the most important features of the map canvas: the ability to change the symbology used for an item based on some data attribute. The simplest example might be to change the line color of a road based on whether its road-type attribute is "highway," "service road," "residential," or so on. But the same feature is used to automatically highlight layer information based on the filtering and querying functionality discussed above. The new release allows many more map attributes to be controlled by these "data definition" settings, and provides a hard-to-miss button next to each attribute, through which a custom data definition can be set.

+

QGIS's composer module is the tool used to take project data and generate a map that can be used outside of the application (in print, as a static image, or as a layer for MapServer or some other software tool, for example). Consequently, it is not a simple select-and-click-export tool; composing the output can involve a lot of choices about which data to make visible, how (and where) to label it, and how to make it generally accessible.

+

The updated composer in 2.8 now has a full-screen mode and sports several new options for configuring output. For instance, the user now has full control over how map axes are labeled. In previous releases, the grid coordinates of the map could be turned on or off, but the only options were all or nothing. Now, the user can individually choose whether coordinates are displayed on all four sides, and can even choose in which direction vertical text labels will run (so that they can be correctly justified to the edge of the map, for example).

+

There are, as usual, many more changes than there is room to discuss. Some particularly noteworthy improvements include the ability to save and load bookmarks for frequently used data sources (perhaps most useful for databases, web services, and other non-local data) and improvements to QGIS's server module. This module allows one QGIS instance to serve up data accessible to other QGIS applications (for example, to simply team projects). The server can now be extended with Python plugins and the data layers that it serves can be styled with style rules like those used in the desktop interface.

+

QGIS is one of those rare free-software applications that is both powerful enough for high-end work and yet also straightforward to use for the simple tasks that might attract a newcomer to GIS in the first place. The 2.8 release, particularly with its project-wide commitment to long-term support, appears to be an update well worth checking out, whether one needs to create a simple, custom map or to mine a database for obscure geo-referenced meaning.

+

Comments (3 posted)

+

Development activity in LibreOffice and OpenOffice

+

By Jonathan Corbet +
March 25, 2015

+

The LibreOffice project was

announced +

with great fanfare in September 2010. Nearly one year later, the OpenOffice.org project (from which LibreOffice was forked)

was cut loose from Oracle -

and found a new home as an Apache project. It is fair to say that the rivalry between the two projects in the time since then has been strong. Predictions that one project or the other would fail have not been borne out, but that does not mean that the two projects are equally successful. A look at the two projects' development communities reveals some interesting differences.

-

Release histories

-

Apache OpenOffice has made two releases in the past year: 4.1 in April 2014 and 4.1.1 (described as "a micro update" in the release announcement) in August. The main feature added during that time would appear to be significantly improved accessibility support.

-

The release history for LibreOffice tells a slightly different story:

-
-

It seems clear that LibreOffice has maintained a rather more frenetic release cadence, generally putting out at least one release per month. The project typically keeps at least two major versions alive at any one time. Most of the releases are of the minor, bug-fix variety, but there have been two major releases in the last year as well.

-

Development statistics

-

In the one-year period since late March 2014, there have been 381 changesets committed to the OpenOffice Subversion repository. The most active committers are:

-
- - - - - - - - - - -
Most active OpenOffice developers
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
By changesets
Herbert Dürr6316.6%
Jürgen Schmidt             5614.7%
Armin Le Grand5614.7%
Oliver-Rainer Wittmann4612.1%
Tsutomu Uchino338.7%
Kay Schenk277.1%
Pedro Giffuni236.1%
Ariel Constenla-Haile225.8%
Andrea Pescetti143.7%
Steve Yin112.9%
Andre Fischer102.6%
Yuri Dario71.8%
Regina Henschel61.6%
Juan C. Sanz20.5%
Clarence Guo20.5%
Tal Daniel20.5%
-
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
By changed lines
Jürgen Schmidt             45549988.1%
Andre Fischer261483.8%
Pedro Giffuni231833.4%
Armin Le Grand110181.6%
Juan C. Sanz45820.7%
Oliver-Rainer Wittmann43090.6%
Andrea Pescetti39080.6%
Herbert Dürr28110.4%
Tsutomu Uchino19910.3%
Ariel Constenla-Haile12580.2%
Steve Yin10100.1%
Kay Schenk6160.1%
Regina Henschel4170.1%
Yuri Dario2680.0%
tal160.0%
Clarence Guo110.0%
-
-
-

In truth, the above list is not just the most active OpenOffice developers — it is all of them; a total of 16 developers have committed changes to OpenOffice in the last year. Those developers changed 528,000 lines of code, but, as can be seen above, Jürgen Schmidt accounted for the bulk of those changes, which were mostly updates to translation files.

-

The top four developers in the "by changesets" column all work for IBM, so IBM is responsible for a minimum of about 60% of the changes to OpenOffice in the last year.

-

The picture for LibreOffice is just a little bit different; in the same one-year period, the project has committed 22,134 changesets from 268 developers. The most active of these developers were:

-
- - - - - - - - - - -
Most active LibreOffice developers
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
By changesets
Caolán McNamara430719.5%
Stephan Bergmann235110.6%
Miklos Vajna14496.5%
Tor Lillqvist11595.2%
Noel Grandin10644.8%
Markus Mohrhard9354.2%
Michael Stahl9154.1%
Kohei Yoshida7553.4%
Tomaž Vajngerl6583.0%
Thomas Arnhold6192.8%
Jan Holesovsky4662.1%
Eike Rathke4572.1%
Matteo Casalin4422.0%
Bjoern Michaelsen4211.9%
Chris Sherlock3961.8%
David Tardon3861.7%
Julien Nabet3621.6%
Zolnai Tamás3381.5%
Matúš Kukan2561.2%
Robert Antoni Buj Gelonch2311.0%
-
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
By changed lines
Lionel Elie Mamane24406212.5%
Noel Grandin23871112.2%
Stephan Bergmann1612208.3%
Miklos Vajna1293256.6%
Caolán McNamara975445.0%
Tomaž Vajngerl694043.6%
Tor Lillqvist594983.1%
Laurent Balland-Poirier528022.7%
Markus Mohrhard505092.6%
Kohei Yoshida455142.3%
Chris Sherlock367881.9%
Peter Foley343051.8%
Christian Lohmaier337871.7%
Thomas Arnhold327221.7%
David Tardon216811.1%
David Ostrovsky216201.1%
Jan Holesovsky207921.1%
Valentin Kettner205261.1%
Robert Antoni Buj Gelonch204471.0%
Michael Stahl182160.9%
-
-
-

To a first approximation, the top ten companies supporting LibreOffice in the last year are:

-
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Companies supporting LibreOffice development
(by changesets)
Red Hat841738.0%
Collabora Multimedia653129.5%
(Unknown)512623.2%
(None)14906.7%
Canonical4221.9%
Igalia S.L.800.4%
Ericsson210.1%
Yandex180.1%
FastMail.FM170.1%
SUSE70.0%
-
-

Development work on LibreOffice is thus concentrated in a small number of companies, though it is rather more spread out than OpenOffice development. It is worth noting that the LibreOffice developers with unknown affiliation, who contributed 23% of the changes, make up 82% of the developer base, so there would appear to be a substantial community of developers contributing from outside the above-listed companies.

-

Some conclusions

-

Last October, some concerns were raised on the OpenOffice list about the health of that project's community. At the time, Rob Weir shrugged them off as the result of a marketing effort by the LibreOffice crowd. There can be no doubt that the war of words between these two projects has gotten tiresome at times, but, looking at the above numbers, it is hard not to conclude that there is an issue that goes beyond marketing hype here.

-

In the 4½ years since its founding, the LibreOffice project has put together a community with over 250 active developers. There is support from multiple companies and an impressive rate of patches going into the project's repository. The project's ability to sustain nearly monthly releases on two branches is a direct result of that community's work. Swearing at LibreOffice is one of your editor's favorite pastimes, but it seems clear that the project is on a solid footing with a healthy community.

-

OpenOffice, instead, is driven by four developers from a single company — a company that appears to have been deemphasizing OpenOffice work for some time. As a result, the project's commit rate is a fraction of what LibreOffice is able to sustain and releases are relatively rare. As of this writing, the OpenOffice +

and found a new home as an Apache project. It is fair to say that the rivalry between the two projects in the time since then has been strong. Predictions that one project or the other would fail have not been borne out, but that does not mean that the two projects are equally successful. A look at the two projects' development communities reveals some interesting differences.

+

Release histories

+

Apache OpenOffice has made two releases in the past year: 4.1 in April 2014 and 4.1.1 (described as "a micro update" in the release announcement) in August. The main feature added during that time would appear to be significantly improved accessibility support.

+

The release history for LibreOffice tells a slightly different story:

+
+

It seems clear that LibreOffice has maintained a rather more frenetic release cadence, generally putting out at least one release per month. The project typically keeps at least two major versions alive at any one time. Most of the releases are of the minor, bug-fix variety, but there have been two major releases in the last year as well.

+

Development statistics

+

In the one-year period since late March 2014, there have been 381 changesets committed to the OpenOffice Subversion repository. The most active committers are:

+
+ + + + + + + + + + +
Most active OpenOffice developers
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
By changesets
Herbert Dürr6316.6%
Jürgen Schmidt             5614.7%
Armin Le Grand5614.7%
Oliver-Rainer Wittmann4612.1%
Tsutomu Uchino338.7%
Kay Schenk277.1%
Pedro Giffuni236.1%
Ariel Constenla-Haile225.8%
Andrea Pescetti143.7%
Steve Yin112.9%
Andre Fischer102.6%
Yuri Dario71.8%
Regina Henschel61.6%
Juan C. Sanz20.5%
Clarence Guo20.5%
Tal Daniel20.5%
+
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
By changed lines
Jürgen Schmidt             45549988.1%
Andre Fischer261483.8%
Pedro Giffuni231833.4%
Armin Le Grand110181.6%
Juan C. Sanz45820.7%
Oliver-Rainer Wittmann43090.6%
Andrea Pescetti39080.6%
Herbert Dürr28110.4%
Tsutomu Uchino19910.3%
Ariel Constenla-Haile12580.2%
Steve Yin10100.1%
Kay Schenk6160.1%
Regina Henschel4170.1%
Yuri Dario2680.0%
tal160.0%
Clarence Guo110.0%
+
+
+

In truth, the above list is not just the most active OpenOffice developers — it is all of them; a total of 16 developers have committed changes to OpenOffice in the last year. Those developers changed 528,000 lines of code, but, as can be seen above, Jürgen Schmidt accounted for the bulk of those changes, which were mostly updates to translation files.

+

The top four developers in the "by changesets" column all work for IBM, so IBM is responsible for a minimum of about 60% of the changes to OpenOffice in the last year.

+

The picture for LibreOffice is just a little bit different; in the same one-year period, the project has committed 22,134 changesets from 268 developers. The most active of these developers were:

+
+ + + + + + + + + + +
Most active LibreOffice developers
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
By changesets
Caolán McNamara430719.5%
Stephan Bergmann235110.6%
Miklos Vajna14496.5%
Tor Lillqvist11595.2%
Noel Grandin10644.8%
Markus Mohrhard9354.2%
Michael Stahl9154.1%
Kohei Yoshida7553.4%
Tomaž Vajngerl6583.0%
Thomas Arnhold6192.8%
Jan Holesovsky4662.1%
Eike Rathke4572.1%
Matteo Casalin4422.0%
Bjoern Michaelsen4211.9%
Chris Sherlock3961.8%
David Tardon3861.7%
Julien Nabet3621.6%
Zolnai Tamás3381.5%
Matúš Kukan2561.2%
Robert Antoni Buj Gelonch2311.0%
+
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
By changed lines
Lionel Elie Mamane24406212.5%
Noel Grandin23871112.2%
Stephan Bergmann1612208.3%
Miklos Vajna1293256.6%
Caolán McNamara975445.0%
Tomaž Vajngerl694043.6%
Tor Lillqvist594983.1%
Laurent Balland-Poirier528022.7%
Markus Mohrhard505092.6%
Kohei Yoshida455142.3%
Chris Sherlock367881.9%
Peter Foley343051.8%
Christian Lohmaier337871.7%
Thomas Arnhold327221.7%
David Tardon216811.1%
David Ostrovsky216201.1%
Jan Holesovsky207921.1%
Valentin Kettner205261.1%
Robert Antoni Buj Gelonch204471.0%
Michael Stahl182160.9%
+
+
+

To a first approximation, the top ten companies supporting LibreOffice in the last year are:

+
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
Companies supporting LibreOffice development
(by changesets)
Red Hat841738.0%
Collabora Multimedia653129.5%
(Unknown)512623.2%
(None)14906.7%
Canonical4221.9%
Igalia S.L.800.4%
Ericsson210.1%
Yandex180.1%
FastMail.FM170.1%
SUSE70.0%
+
+

Development work on LibreOffice is thus concentrated in a small number of companies, though it is rather more spread out than OpenOffice development. It is worth noting that the LibreOffice developers with unknown affiliation, who contributed 23% of the changes, make up 82% of the developer base, so there would appear to be a substantial community of developers contributing from outside the above-listed companies.

+

Some conclusions

+

Last October, some concerns were raised on the OpenOffice list about the health of that project's community. At the time, Rob Weir shrugged them off as the result of a marketing effort by the LibreOffice crowd. There can be no doubt that the war of words between these two projects has gotten tiresome at times, but, looking at the above numbers, it is hard not to conclude that there is an issue that goes beyond marketing hype here.

+

In the 4½ years since its founding, the LibreOffice project has put together a community with over 250 active developers. There is support from multiple companies and an impressive rate of patches going into the project's repository. The project's ability to sustain nearly monthly releases on two branches is a direct result of that community's work. Swearing at LibreOffice is one of your editor's favorite pastimes, but it seems clear that the project is on a solid footing with a healthy community.

+

OpenOffice, instead, is driven by four developers from a single company — a company that appears to have been deemphasizing OpenOffice work for some time. As a result, the project's commit rate is a fraction of what LibreOffice is able to sustain and releases are relatively rare. As of this writing, the OpenOffice blog shows no posts in 2015. In the October discussion, Rob said that "the dogs may bark but the caravan moves on." That may be true, but, in this case, the caravan does not appear to be moving with any great speed.

-

Anything can happen in the free-software development world; it is entirely possible that a reinvigorated OpenOffice.org may yet give LibreOffice a run for its money. But something will clearly have to change to bring that future around. As things stand now, it is hard not to conclude that LibreOffice has won the battle for developer participation.

-

Comments (74 posted)

-

Page editor: Jonathan Corbet -

-

Inside this week's LWN.net Weekly Edition

-
    -
  • Security: Toward secure package downloads; New vulnerabilities in drupal, mozilla, openssl, python-django ...
  • -
  • Kernel: LSFMM coverage: NFS, defragmentation, epoll(), copy offload, and more.
  • -
  • Distributions: A look at Debian's 2015 DPL candidates; Debian, Fedora, ...
  • -
  • Development: A look at GlusterFS; LibreOffice Online; Open sourcing existing code; Secure Boot in Windows 10; ...
  • -
  • Announcements: A Turing award for Michael Stonebraker, Sébastien Jodogne, ReGlue are Free Software Award winners, Kat Walsh joins FSF board of directors, Cyanogen, ...
  • -
Next page -

:

Security>> -
-
- - +

Anything can happen in the free-software development world; it is entirely possible that a reinvigorated OpenOffice.org may yet give LibreOffice a run for its money. But something will clearly have to change to bring that future around. As things stand now, it is hard not to conclude that LibreOffice has won the battle for developer participation.

+

Comments (74 posted)

+

Page editor: Jonathan Corbet +

+

Inside this week's LWN.net Weekly Edition

+
    +
  • Security: Toward secure package downloads; New vulnerabilities in drupal, mozilla, openssl, python-django ...
  • +
  • Kernel: LSFMM coverage: NFS, defragmentation, epoll(), copy offload, and more.
  • +
  • Distributions: A look at Debian's 2015 DPL candidates; Debian, Fedora, ...
  • +
  • Development: A look at GlusterFS; LibreOffice Online; Open sourcing existing code; Secure Boot in Windows 10; ...
  • +
  • Announcements: A Turing award for Michael Stonebraker, Sébastien Jodogne, ReGlue are Free Software Award winners, Kat Walsh joins FSF board of directors, Cyanogen, ...
  • +
Next page +

:

Security>> +
\ No newline at end of file