John Nilsson wrote:
> Hi,
> 
> Are the dwellers of this list involved in (or otherwise have opinions
> about) the Pirate Bays attempt to engineer a BitTorrent replacement?[1]

I've not been following closely.  But I seems like the mainline bit-torrent
folks are going closed source, where you can apply and they might let you see 
the source... IMO clearly unhealthy.

A next generation client should have an opensource client that demonstrated
the new functionality, define a standard for the DHT, and the protocol
(with room for extensions of course).

Streaming is an obvious goal, and folks are already trying to do it (tribler
and http://distribustream.org/ among others.)  Not sure if the next generation
torrent should cater to the closed tracker folks or not, if I had my druthers
they wouldn't use trackers at all.  Kudos for distribustream for being GPL and
tribler for being mostly GPL (I've not reviewed the various other licenses 
mentioned for tribler).

Support for a sloppy DHT seems like a big win, dramatically better performance 
for popular queries.  Seems that with millions (1.1M on azureus and compatible 
clients at the moment) that the performance wins could be significant.  Google
for sloppy DHT if you want to read a paper on it, good paper, it's the first
hit.

Ideally a faster DHT would make it easier for clients to find network local
clients.  Nothing like being on a big swarm behind a limited downlink and not
finding other peers behind the same downlink (yes I know about various
plugins or adaptations for discovery and manually adding peers).

Significantly faster DHT might allow for other yet to be developed applications.

Ideally the next generation replacement would figure out the bits required
to let a healthy ecosystem of clients provide streaming, discovery of content,
reputation and the like.  A switch to sha256 checksums for verification and
support of signed publishers would be ideal.

Torrents basically work, I think someone else needs to replace the mainline
bit torrent folks and allow for a slow evolution.  Code should be GPL
_______________________________________________
p2p-hackers mailing list
[email protected]
http://lists.zooko.com/mailman/listinfo/p2p-hackers

Reply via email to