[BitTorrent] extending the tracker

silvan zurbruegg silvan at bitflux.ch
Mon Oct 11 00:34:11 EDT 2004

Hi All

I had this idea about extending the tracker, so
that it doesn't just negotiates downloaders but
also knows about a file from the .torrent.

The idea is basically to let the tracker store
informations about a file in some database. 
It should be capable of handling torrent informations
on upload, update and for queries about a file.

>From skimming through the code i basically considered
the following attempts;

- Mapping a request path in track.py to requests on torrent files
  (eg. http://trackerurl:6969/metainfo/).
  The client sends metainformations about a file
  as get-params to the according url  

- Catching the event get-param on `started'.
  The client sends metainformations as additional
  get-params to the according url. The tracker can
  check (with the info hash) whether he knows a file
  and inserts/updates the metainformations.

- A client can query a tracker about specific files.
- A tracker can generate a website on the fly containing
  metainformations and dyn. generated .torrents.

Please let me know your opinion about it or whether there
were already attempts going that direction etc.

Thanks in advance

------------------------ Yahoo! Groups Sponsor --------------------~--> 
$9.95 domain names from Yahoo!. Register anything.

Yahoo! Groups Links

<*> To visit your group on the web, go to:

<*> To unsubscribe from this group, send an email to:
    BitTorrent-unsubscribe at yahoogroups.com

<*> Your use of Yahoo! Groups is subject to:

More information about the BitTorrent mailing list