全部博文（24）
分类： 系统运维
2009-05-18 10:24:28
is a peer-to-peer file sharing protocol designed by Bram Cohen. Visit his pages at BitTorrent is designed to facilitate file transfers among multiple peers across unreliable networks.
The purpose of this specification is to document version 1.0 of the BitTorrent protocol specification in detail. Bram's outlines the protocol in somewhat general terms, and lacks behaviorial detail in some areas. The hope is that this document will become a formal specification, written in clear, unambiguous terms, which can be used as a basis for discussion and implementation in the future.
This document is intended to be maintained and used by the BitTorrent development community. Everyone is invited to contribute to this document, with the understanding that the content here is intended to represent the current protocol, which is already deployed in a number of existing client implementations.
This is not the place to suggest feature requests. For that, please go to the .
This document applies to the first version (i.e. version 1.0) of the
BitTorrent protocol specification. Currently, this applies to the
torrent file structure, peer wire protocol, and the Tracker HTTP/HTTPS
protocol specifications. As newer revisions of each protocol are
defined, they should be specified on their own separate pages, not here.
In this document, a number of conventions are used in an attempt to present information in a concise and unambiguous fashion.
In order to help others find recent changes that have been made to this document, please fill out the change log (last section). This should contain a brief (i.e. one-line) entry for each major change that you've made to the document.
Bencoding is a way to specify and organize data in a terse format. It supports the following types: byte strings, integers, lists, and dictionaries.
Byte strings are encoded as follows:
Note that there is no constant beginning delimiter, and no ending delimiter.
Integers are encoded as follows:
The initial i and trailing e are beginning and ending delimiters. You can have negative numbers such as i-3e. You cannot prefix the number with a zero such as i04e. However, i0e is valid.
Lists are encoded as follows:
The initial l and trailing e are beginning and ending delimiters. Lists may contain any bencoded type, including integers, strings, dictionaries, and other lists.
Dictionaries are encoded as follows:
The initial d and trailing e are the beginning and ending delimiters. Note that the keys must be bencoded strings. The values may be any bencoded type, including integers, strings, lists, and other dictionaries. Keys must be strings and appear in sorted order (sorted as raw strings, not alphanumerics). The strings should be compared using a binary comparison, not a culture-specific "natural" comparison.
All data in a metainfo file is bencoded. The specification for bencoding is defined above.
The content of a metainfo file (the file ending in ".torrent") is a bencoded dictionary, containing the keys listed below. All character string values are UTF-8 encoded.
This section contains the field which are common to both mode, "single file" and "multiple file".
For the case of the single-file mode, the info dictionary contains the following structure:
For the case of the multi-file mode, the info dictionary contains the following structure:
The tracker is an HTTP/HTTPS service which responds to HTTP GET requests. The requests include metrics from clients that help the tracker keep overall statistics about the torrent. The response includes a peer list that helps the client participate in the torrent. The base URL consists of the "announce URL" as defined in the metadata (.torrent) file. The parameters are then added to this URL, using standard CGI methods (i.e. a '?' after the announce URL, followed by 'param=value' sequences separated by '&').
Note that all binary data in the URL (particularly info_hash and peer_id) must be properly escaped. This means any byte not in the set 0-9, a-z, A-Z, '.', '-', '_' and '~', must be encoded using the "%nn" format, where nn is the hexadecimal value of the byte. (See for details.)
For a 20-byte hash of \x12\x34\x56\x78\x9a\xbc\xde\xf1\x23\x45\x67\x89\xab\xcd\xef\x12\x34\x56\x78\x9a,
The right encoded form is %124Vx%9A%BC%DE%F1%23Eg%89%AB%CD%EF%124Vx%9A
The parameters used in the client->tracker GET request are as follows:
The tracker responds with "text/plain" document consisting of a bencoded dictionary with the following keys:
As mentioned above, the list of peers is length 50 by default. If there are fewer peers in the torrent, then the list will be smaller. Otherwise, the tracker randomly selects peers to include in the response. The tracker may choose to implement a more intelligent mechanism for peer selection when responding to a request. For instance, reporting seeds to other seeders could be avoided.
Clients may send a request to the tracker more often than the specified interval, if an event occurs (i.e. stopped or completed) or if the client needs to learn about more peers. However, it is considered bad practice to "hammer" on a tracker to get multiple peers. If a client wants a large peer list in the response, then it should specify the numwant parameter.
Implementer's Note: Even 30 peers is plenty, the official client version 3 in fact only actively forms new connections if it has less than 30 peers and will refuse connections if it has 55. This value is important to performance. When a new piece has completed download, HAVE messages (see below) will need to be sent to most active peers. As a result the cost of broadcast traffic grows in direct proportion to the number of peers. Above 25, new peers are highly unlikely to increase download speed. UI designers are strongly advised to make this obscure and hard to change as it is very rare to be useful to do so.
By convention most trackers support another form of request, which queries the state of a given torrent (or all torrents) that the tracker is managing. This is referred to as the "scrape page" because it automates the otherwise tedious process of "screen scraping" the tracker's stats page.
The scrape URL is also a HTTP GET method, similar to the one described above. However the base URL is different. To derive the scrape URL use the following steps: Begin with the announce URL. Find the last '/' in it. If the text immediately following that '/' isn't 'announce' it will be taken as a sign that that tracker doesn't support the scrape convention. If it does, substitute 'scrape' for 'announce' to find the scrape page.
Examples: (announce URL -> scrape URL)
~ -> ~
~ -> ~
~.php -> ~.php
~ -> (scrape not supported)
~?x2%0644 -> ~?x2%0644
~?x=2/4 -> (scrape not supported)
~%064announce -> (scrape not supported)
Note especially that entity unquoting is not to be done. This standard is documented by Bram in the development list archive:
The scrape URL may be supplemented by the optional parameter info_hash, a 20-byte value as described above. This restricts the tracker's report to that particular torrent. Otherwise stats for all torrents that the tracker is managing are returned. Software authors are strongly encouraged to use the info_hash parameter when at all possible, to reduce the load and bandwidth of the tracker.
You may also specify multiple info_hash parameters to trackers that support it. While this isn't part of the official specifications it has become somewhat a defacto standard - for example:
.php?info_hash=aaaaaaaaaaaaaaaaaaaa&info_hash=bbbbbbbbbbbbbbbbbbbb&info_hash=cccccccccccccccccccc
The response of this HTTP GET method is a "text/plain" or sometimes gzip compressed document consisting of a bencoded dictionary, containing the following keys:
Note that this response has three levels of dictionary nesting. Here's an example:
d5:filesd20:....................d8:completei5e10:downloadedi50e10:incompletei10eeee
Where .................... is the 20 byte info_hash and there are 5 seeders, 10 leechers, and 50 complete downloads.
Below are the response keys are being unofficially used. Since they are unofficial, they are all optional.
The peer protocol facilitates the exchange of pieces as described in the 'metainfo file.
Note here that the original specification also used the term "piece" when describing the peer protocol, but as a different term than "piece" in the metainfo file. For that reason, the term "block" will be used in this specification to describe the data that is exchanged between peers over the wire.
A client must maintain state information for each connection that it has with a remote peer:
Note that this also implies that the client will also need to keep track of whether or not it is interested in the remote peer, and if it has the remote peer choked or unchoked. So, the real list looks something like this:
Client connections start out as "choked" and "not interested". In other words:
A block is downloaded by the client when the client is interested in a peer, and that peer is not choking the client. A block is uploaded by a client when the client is not choking a peer, and that peer is interested in the client.
It is important for the client to keep its peers informed as to whether or not it is interested in them. This state information should be kept up-to-date with each peer even when the client is choked. This will allow peers to know if the client will begin downloading when it is unchoked (and vice-versa).
Unless specified otherwise, all integers in the peer wire protocol are encoded as four byte big-endian values. This includes the length prefix on all messages that come after the handshake.
The peer wire protocol consists of an initial handshake. After that, peers communicate via an exchange of length-prefixed messages. The length-prefix is an integer as described above.
The handshake is a required message and must be the first message transmitted by the client. It is (49+len(pstr)) bytes long.
handshake:
In version 1.0 of the BitTorrent protocol, pstrlen = 19, and pstr = "BitTorrent protocol".
The initiator of a connection is expected to transmit their handshake immediately. The recipient may wait for the initiator's handshake, if it is capable of serving multiple torrents simultaneously (torrents are uniquely identified by their infohash). However, the recipient must respond as soon as it sees the info_hash part of the handshake. The tracker's NAT-checking feature does not send the peer_id field of the handshake.
If a client receives a handshake with an info_hash that it is not currently serving, then the client must drop the connection.
If the initiator of the connection receives a handshake in which the peer_id does not match the expected peerid, then the initiator is expected to drop the connection. Note that the initiator presumably received the peer information from the tracker, which includes the peer_id that was registered by the peer. The peer_id from the tracker and in the handshake are expected to match.
The peer_id is exactly 20 bytes (characters) long.
There are mainly two conventions how to encode client and client version information into the peer_id, Azureus-style and Shadow's-style.
Azureus-style uses the following encoding: '-', two characters for client id, four ascii digits for version number, '-', followed by random numbers.
For example: '-AZ2060-'...
known clients that uses this encoding style are:
Clients which have been seen in the wild and need to be identified:
Shadow's style uses the following encoding: one ascii alphanumeric for client identification, up to five characters for version number (padded with '-' if less than five), followed by three characters (commonly '---', but not always the case), followed by random characters. Each character in the version string represents a number from 0 to 63. '0'=0, ..., '9'=9, 'A'=10, ..., 'Z'=35, 'a'=36, ..., 'z'=61, '.'=62, '-'=63.
A full explanation by Shad0w about the encoding style (including information about existing conventions on how the three characters after the version string are used) can be found .
For example: 'S58B-----'... for Shadow's 5.8.11
known clients that uses this encoding style are:
Bram's client now uses this style... 'M3-4-2--' or 'M4-20-8-'.
does something different still. Its peer_id consists of four ASCII characters 'exbc', followed by two bytes x and y, followed by random characters. The version number is x in decimal before the decimal point and y as two decimal digits after the decimal point. uses the same scheme, but adds 'LORD' after the version bytes. An for BitComet once replaced 'exbc' with 'FUTB'. The encoding for BitComet Peer IDs changed to Azureus-style as of BitComet version 0.59.
has its own style too. Its peer_id consists of the three uppercase characters 'XBT' followed by three ASCII digits representing the version number. If the client is a debug build, the seventh byte is the lowercase character 'd', otherwise it is a '-'. Following that is a '-' then random digits, uppercase and lowercase letters. Example: 'XBT054d-' at the beginning would indicate a debug build of version 0.5.4.
use the following peer_id scheme: The first two characters are 'OP' and the next four digits equal the build number. All following characters are random lowercase hexdecimal digits.
use the following peer_id scheme: the first characters are '-ML' followed by a dotted version then a '-' followed by randomness. e.g. '-ML2.7.2-kgjjfkd'
uses the pattern '-BOWxxx-yyyyyyyyyyyy', where y is random (uppercase letters) and x depends on the version. Version 1.0.6 has xxx = A0C.
uses Bram's new style: 'Q1-0-0--' or 'Q1-10-0-' followed by random bytes.
is an Azureus fork and simply uses 'AZ2500BT' + random bytes as peer ID in its 1.1 version. Note the missing dashes.
version 1.90 pretends to be or is derived from Mainline 3.4.6. Its peer ID starts with '346------'.
has several modes for its peer ID. In one mode it reads the ID of its peer and reconnects using the first eight bytes as a basis for its own ID. Its real ID appears to use '\0\3BS' (C notation) as the first four bytes for version 3.x and '\0\2BS' for version 2.x. In all modes the ID may end in 'UDP0'.
uses its version as decimal ASCII values for the first two bytes. The third and fourth bytes are 'RS'. What then follows is the nickname of the user and some random bytes.
starts its peer ID with '-G3' and appends up to 9 characters of the nickname of the user.
uses Azureus style with 'FG' but without the trailing '-'. Version 1.82.1002 still uses the version digits '0180'.
is derived from BitTornado but tries to mimic Azureus style. The result is that its peer ID starts with '-NE', continues with a 4 digit version number and then directly goes on with the three characters that describe the type of client in Shad0w's peer ID style.
takes the sha1 hash of a user dependent string and replaces the first few characters with "AP" + version string + "-".
starts its id with the four letters "QVOD" and continues with its build number in four decimal digits (currently "0054"). The remaining 12 characters are random uppercase hexdecimal digits. There appears to be a popular modified client in China that replaces the four characters in the beginning with random bytes.
Many clients are using all random numbers or 12 zeroes followed by random numbers (like older versions of ).
All of the remaining messages in the protocol take the form of
The keep-alive message is a message with zero bytes, specified with the length prefix set to zero. There is no message ID and no payload. Peers may close a connection if they receive no messages (keep-alive or any other message) for a certain period of time, so a keep-alive message must be sent to maintain the connection alive if no command have been sent for a given amount of time. This amount of time is generally two minutes.
The choke message is fixed-length and has no payload.
The unchoke message is fixed-length and has no payload.
The interested message is fixed-length and has no payload.
The not interested message is fixed-length and has no payload.
The have message is fixed length. The payload is the zero-based index of a piece that has just been successfully downloaded and verified via the hash.
Implementer's Note: That is the strict definition, in reality some games may be played. In particular because peers are extremely unlikely to download pieces that they already have, a peer may choose not to advertise having a piece to a peer that already has that piece. At a minimum "HAVE supression" will result in a 50% reduction in the number of HAVE messages, this translates to around a 25-35% reduction in protocol overhead. At the same time, it may be worthwhile to send a HAVE message to a peer that has that piece already since it will be useful in determining which piece is rare.
A malicious peer might also choose to advertise having pieces that it knows the peer will never download. Due to this attempting to model peers using this information is a bad idea.
The bitfield message may only be sent immediately after the handshaking sequence is completed, and before any other messages are sent. It is optional, and need not be sent if a client has no pieces.
The bitfield message is variable length, where X is the length of the bitfield. The payload is a bitfield representing the pieces that have been successfully downloaded. The high bit in the first byte corresponds to piece index 0. Bits that are cleared indicated a missing piece, and set bits indicate a valid and available piece. Spare bits at the end are set to zero.
A bitfield of the wrong length is considered an error. Clients should drop the connection if they receive bitfields that are not of the correct size, or if the bitfield has any of the spare bits set.
The request message is fixed length, and is used to request a block. The payload contains the following information:
This section is under dispute! Please use the discussion page to resolve this!
View #1 According to the official specification, "All current implementations use 2^15 (32KB), and close connections which request an amount greater than 2^17 (128KB)." As early as version 3 or 2004, this behavior was changed to use 2^14 (16KB) blocks. As of version 4.0 or mid-2005, the mainline disconnected on requests larger than 2^14 (16KB); and some clients have followed suit. Note that block requests are smaller than pieces (>=2^18 bytes), so multiple requests will be needed to download a whole piece.
Strictly, the specification allows 2^15 (32KB) requests. The reality is near all clients will now use 2^14 (16KB) requests. Due to clients that enforce that size, it is recommended that implementations make requests of that size. Due to smaller requests resulting in higher overhead due to tracking a greater number of requests, implementers are advised against going below 2^14 (16KB).
The choice of request block size limit enforcement is not nearly so clear cut. With mainline version 4 enforcing 16KB requests, most clients will use that size. At the same time 2^14 (16KB) is the semi-official (only semi because the official protocol document has not been updated) limit now, so enforcing that isn't wrong. At the same time, allowing larger requests enlarges the set of possible peers, and except on very low bandwidth connections (<256kbps) multiple blocks will be downloaded in one choke-timeperiod, thus merely enforcing the old limit causes minimal performance degradation. Due to this factor, it is recommended that only the older 2^17 (128KB) maximum size limit be enforced.
View #2 This section has contained falsehoods for a large portion of the time this page has existed. This is the third time I (uau) am correcting this same section for incorrect information being added, so I won't rewrite it completely since it'll probably be broken again... Current version has at least the following errors: Mainline started using 2^14 (16384) byte requests when it was still the only client in existence; only the "official specification" still talked about the obsolete 32768 byte value which was in reality neither the default size nor maximum allowed. In version 4 the request behavior did not change, but the maximum allowed size did change to equal the default size. In latest mainline versions the max has changed to 32768 (note that this is the first appearance of 32768 for either default or max size since the first ancient versions). "Most older clients use 32KB requests" is false. Discussion of larger requests fails to take latency effects into account.
The piece message is variable length, where X is the length of the block. The payload contains the following information:
The cancel message is fixed length, and is used to cancel block requests. The payload is identical to that of the "request" message. It is typically used during "End Game" (see the Algorithms section below).
The port message is sent by newer versions of the Mainline
that implements a DHT tracker. The listen port is the port this peer's
DHT node is listening on. This peer should be inserted in the local
routing table (if DHT tracker is supported).
This section is under dispute! Please use the discussion page to resolve this!
View #1In general peers are advised to keep a few unfullfilled requests on each connection. This is done because otherwise a full round trip is required from the download of one block to begining the download of a new block (round trip between PIECE message and next REQUEST message). On links with high BDP (bandwidth-delay-product, high latency or high bandwidth), this can result in a substantial performance loss.
Implementer's note: This the 'most crucial performance item. A static queue of 10 requests is reasonable for 16KB blocks on a 5mbps link with 50ms latency. Links with greater bandwidth are becoming very common so UI designers are urged to make this readily available for changing. Notably cable modems were known for traffic policing and increasing this might of aleviated some of the problems caused by this.
View #2 NOTE: much of the information in this "Queuing" section is false or misleading. I'll just note that the "defaults to 5 outstanding requests" hasn't been true for a long time, "32 KB blocks" is misleading since you normally don't use 32 KB blocks, and tuning queue length by changing it and trying to measure the effects is a bad idea.
(This was not part of the original specification)
The super-seed feature in S-5.5 and on is a new seeding algorithm designed to help a torrent initiator with limited bandwidth "pump up" a large torrent, reducing the amount of data it needs to upload in order to spawn new seeds in the torrent.
When a seeding client enters "super-seed mode", it will not act as a standard seed, but masquerades as a normal client with no data. As clients connect, it will then inform them that it received a piece -- a piece that was never sent, or if all pieces were already sent, is very rare. This will induce the client to attempt to download only that piece.
When the client has finished downloading the piece, the seed will not inform it of any other pieces until it has seen the piece it had sent previously present on at least one other client. Until then, the client will not have access to any of the other pieces of the seed, and therefore will not waste the seed's bandwidth.
This method has resulted in much higher seeding efficiencies, by both inducing peers into taking only the rarest data, reducing the amount of redundant data sent, and limiting the amount of data sent to peers which do not contribute to the swarm. Prior to this, a seed might have to upload 150% to 200% of the total size of a torrent before other clients became seeds. However, a large torrent seeded with a single client running in super-seed mode was able to do so after only uploading 105% of the data. This is 150-200% more efficient than when using a standard seed.
Super-seed mode is 'NOT recommended for general use. While it does assist in the wider distribution of rare data, because it limits the selection of pieces a client can downlad, it also limits the ability of those clients to download data for pieces they have already partially retrieved. Therefore, super-seed mode is only recommended for initial seeding servers.
Why not rename it to e.g. "Initial Seeding Mode" or "Releaser Mode" then?
Clients may choose to download pieces in random order.
A better strategy is to download pieces in order. The client can determine this by keeping the initial bitfield from each peer, and updating it with every 'have message. Then, the client can download the pieces that appear least frequently in these peer bitfields. Note that any Rarest First strategy should include randomization among at least several of the least common pieces, as having many clients all attempting to jump on the same "least common" piece would be counter productive.
When a download is almost complete, there's a tendency for the last few blocks to trickle in slowly. To speed this up, the client sends requests for all of its missing blocks to all of its peers. To keep this from becoming horribly inefficient, the client also sends a cancel to everyone else every time a block arrives.
There is no documented thresholds, recommended percentages, or block counts that could be used as a guide or Recommended Best Practice here.
When to enter end game mode is an area of discussion. Some clients enter end game when all pieces have been requested. Others wait until the number of blocks left is lower than the number of blocks in transit, and no more than 20. There seems to be agreement that it's a good idea to keep the number of pending blocks low (1 or 2 blocks) to minimize the overhead, and if you randomize the blocks requested, there's a lower chance of downloading duplicates. More on the protocol overhead can be found here:
Choking is done for several reasons. TCP congestion control behaves very poorly when sending over many connections at once. Also, choking lets each peer use a tit-for-tat-ish algorithm to ensure that they get a consistent download rate.
The choking algorithm described below is the currently deployed one. It is very important that all new algorithms work well both in a network consisting entirely of themselves and in a network consisting mostly of this one.
There are several criteria a good choking algorithm should meet. It should cap the number of simultaneous uploads for good TCP performance. It should avoid choking and unchoking quickly, known as 'fibrillation'. It should reciprocate to peers who let it download. Finally, it should try out unused connections once in a while to find out if they might be better than the currently used ones, known as optimistic unchoking.
The currently deployed choking algorithm avoids fibrillation by only changing choked peers once every ten seconds.
Reciprocation and number of uploads capping is managed by unchoking the four peers which have the best upload rate and are interested. This maximizes the client's download rate. These four peers are referred to as downloaders, because they are interested in downloading from the client.
Peers which have a better upload rate (as compared to the downloaders) but aren't interested get unchoked. If they become interested, the downloader with the worst upload rate gets choked. If a client has a complete file, it uses its upload rate rather than its download rate to decide which peers to unchoke.
For optimistic unchoking, at any one time there is a single peer which is unchoked regardless of its upload rate (if interested, it counts as one of the four allowed downloaders). Which peer is optimistically unchoked rotates every 30 seconds. Newly connected peers are three times as likely to start as the current optimistic unchoke as anywhere else in the rotation. This gives them a decent chance of getting a complete piece to upload.
Occasionally a
peer will be choked by all peers which it was formerly downloading
from. In such cases it will usually continue to get poor download rates
until the optimistic unchoke finds better peers. To mitigate this
problem, when over a minute goes by without getting any piece data
while downloading from a peer,
assumes it is "snubbed" by that peer and doesn't upload to it except as
an optimistic unchoke. This frequently results in more than one
concurrent optimistic unchoke, (an exception to the exactly one
optimistic unchoke rule mentioned above), which causes download rates
to recover much more quickly when they falter.
Currently there are a few official extensions to the protocol.
These extensions serve multiple purposes. They allow a peer to more quickly bootstrap into a swarm by giving a peer a specific set of pieces which they will be allowed download regardless of choked status. They reduce message overhead by adding HaveAll and HaveNone messages and allow explicit rejection of piece requests whereas previously only implicit rejection was possible meaning that a peer might be left waiting for a piece that would never be delivered.
The specificication is documented at the site here:
This extension is to allow for the tracking of peers downloading torrents without the use of a standard tracker. A peer implementing this protocol becomes a "tracker" and stores lists of other nodes/peers which can be used to locate new peers.
The specification is documented at the site here:
This extension allows the creation of encrypted connections between peers. This can be used to bypass ISPs throttling BitTorrent traffic.
The specification is documented at
The documentation is fairly complete, but ideally it would
be clarified on several points including guidance on when encrypted
connections should be attempted, fallback procedures to regular
connections etc.
A protocol in its own right - if two clients indicate they support the protocol, then they should switch over to using it. It allows normal BitTorrent as well extension messages to be sent over it, and is documented . Currently implemented by Azureus and Transmission.
It is not possible to use both this protocol and the LibTorrent extension protocol at the same time - if both clients indicate they support both, then they should follow the semantics defined by the .
The possibility to seed a torrent via a web server is generally called WebSeeding. It allows the HTTP server to work as a peer in the BitTorrent network.
There are at least two specification for how to combine a torrent download with a HTTP download. The first standard, implemented by BitTornado is quite easy to implement in the client, but is intrusive on the HTTP in that it requires a script handling requests on the server side. i.e. A plain HTTP server that just serves plain files isn't enough. The benfits is that the script can be more abuse resistant. This specification is found here:
The second specification requires slightly more from the client, but downloads from plain HTTP servers. It is specified here: . It has been implemented by GetRight, libtorrent and Mainline.
This is a protocol for exchanging extension information and was derived from an early version of azureus' extension protocol. It adds one message for exchanging arbitrary handshake information including defined extension messages, mapping extensions to specific message IDs. It is documented here: and is implemented by libtorrent, uTorrent and Mainline.
It is not possible to use both this protocol and the Azureus Messaging Protocol at the same time - if both clients indicate they support both, then they should follow the semantics defined by the .
These bits are used to allow two clients that support both the Azureus Messaging Protocol and LibTorrent's extension protocol to decide which of the two extensions should be used for communication, and is defined .
A Protocol, considering peers location (in geographical terms) for better performance. Specification can be found .
An extension using message id 9 to add peer exchange and connection statistics exchange. The specification can be found . The extension was in use in SimpleBT 0.32 to 0.36.1. Later versions of SimpleBT were called BitComet and used the similar but incompatible BitComet Extension Protocol.
There appears to be no official documentation.
In this protocol a peer announces the supported extensions by
sending a message
Extensions currently in use (TODO: reverse engineer semantics):
A minimum implementation needs only accept EXT_SUPPORT, but EXT_PEERREQ and EXT_PEERS are supported by all known implementations.
The reserved bits are numbered 1-64 in the following table for ease of identification. Bit 1 corresponds to the most significant bit of the first reserved byte. Bit 8 corresponds to the least significant bit of the first reserved byte (i.e. byte[0] |= 0x01). Bit 64 is the least significant bit of the last reserved byte i.e. byte[7] |= 0x01
An orange bit is a known unofficial extension, a red bit is an unknown unofficial extension.
| Bit | Use | Azureus | BitComet | MainLine | MonoTorrent | µTorrent | libtorrent | KTorrent | BitLord | XBT | Transmission |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Azureus Extended Messaging | Yes | ? | ? | ? | ? | No | No | No | No | Yes |
| 1-16 | BitComet Extension protocol | No | Yes | No | No | No | No | No | Yes | No | No |
| 21 | BitTorrent Location-aware Protocol 1.0 | No | No | No | No | No | No | No | No | No | No |
| 44 | Extension protocol | Yes | ? | Yes | Yes | Yes | Yes | No | No | Yes | Yes |
| 47 - 48 | Extension Negotiation Protocol | Yes | No | No | No | No | No | No | No | No | Yes |
| 61 | NAT Traversal | No | ? | Yes | ? | ? | No | ? | ? | No | ? |
| 62 | Fast Peers | No | ? | Yes | Yes | ? | Yes | Yes | No | No | ? |
| 63 | XBT Peer Exchange | No | ? | No | ? | ? | No | ? | ? | Yes | ? |
| 64 | DHT | No | ? | Yes | Yes | Yes | Yes | Yes | No | No | ? |
| 64 | XBT Metadata Exchange | No | ? | No | ? | ? | No | ? | ? | Yes | ? |
