Hi all,

Do you think there would be ANY adavantage of augmenting gnutella with a push based information flow.

I mean, let us assume that even amongst peerswe are likely to find client-like and server-like peers. i.e, a clientlike peer would conume more data than it would share and the server like peer would share more data than it actually consumes.

I such a situation, let us say, there is a hot data file present with jus one/ or mebbe just a few of the nodes in the network and the presence of the file is not yet known to otherr users.

So the idea is to push these hot data items accross the networks(amongst server like peers) even before the requests for the data arises.


lemme know what you think about the idea,

sara