[p2p-hackers] network coding put to practical p2p content dist. use

[coderman]

On 6/19/05, Nick Johnson <arachnid at notdot.net> wrote:
> ... there's still a
> chance that your peer already has the piece you want to send, but the
> higher the number of pieces, the lower that chance.

in the bittorrent context you can centrally direct the range of blocks
to be xfer'd between peers efficiently regardless of whether the
blocks are error coded or not.  i don't see Bram's arguments intended
for all p2p / decentralized systems [and the myraid ways error coding
can be integrated].


> That said, I'm working on a P2P system where we do intend to use
> Online codes (a variant of Fountain codes) - not for piece encoding,
> but for the transmission of individual pieces via multicast. This is
> where they really shine.

wireless broadcast (at the data packet layer, not just
physical/transport) is another environment where these codes are
useful for the same kind of unidirectional bulk transport.  there are
lots of potential uses for these codes but a centralized tracker is
not one of them.

[it would be interesting to see if Bram reconsiders these type of
decentralized network codes for a trackerless DHT torrent network,
assuming CPU requirements were not prohibitive]