status

in development.  don't expect things to work yet.

intro

this package contains a bittorrent peer (for downloading and uploading
files), tools for verifying torrents & printing information about
torrents, a bittorrent tracker for use with a http server, and a
program for creating torrent files.

install

this package depends on "http", "web" and "util".

to install, bind/symlink/include a mkconfig (e.g. from your inferno
install), and:

mk install

when building from within inferno, insert SYSHOST=Inferno and ROOT=
in the mk invocations to override the values in the mkconfig.

latest

the latest version can be found at

    http://www.ueber.net/code/r/torrent

licence & author

all files are in the public domain.  this code has been written by
mechiel lukkien, reachable at mechiel@ueber.net or mechiel@xs4all.nl.

todo

wm/torrent:

• is there a race in reading Qpeers? e.g. that an old read is happening after a seek 0 for restarting? e.g. due to a slow kill? i've seen an error about a bad line from Qpeers...
• available bar is not smoothly/correctly filled. probably best to only show non-seeders in it. availbar seems to pieces with overlap, on the edges of the piece. figure out how to fill the panel with the bars (and thus bars themselves) in x direction. catch resizes and recreate the bars with new width.
• make torrent/peer send events on removal of bad peers? and remove them from our list. or at least reserve the peerevent message for it. clear stale entries from badpeers list.
• could show dns names instead of/along ip addresses.

torrent/peer

• once all mechanisms are implemented and features settle down, make all the accounting less cpu intensive. torrent/peer takes lots of cpu now.

• allow downloading of a subset of files, perhaps with priority
• implement rarest-first scheme?
• handle multiple torrent files at once? allows for better use of bandwidth
• support magnet uri's? (bep 9)
• udp tracker support (bep 15)
• http seeding (bep 17 and/or bep 19)

torrent/track

• make search for a peer (given ip & port) less inefficient.
• might add scraping.
• might be a bit smarter about which peers to send to a peer. e.g. not itself, make randomizing cheaper, let peers that are done (left=0) not connect to each other.