Servent responding to a PING

S

spooning leads to forking

New member
Hi!

For efficiency reasons why doesn't, in a future version of the protocol, the servent send just one PONG to it's neighbour with all the PONGS he received, or even, with the new cache PONG version that limewire proposed, why not sending just one PONG instead of n ?

Thanks,
TAS
 
Well, pong traffic is not really a problem at the moment and the pong messages serves more purposes than just exploring the size of the reachable network. - LimeWire, for example, uses pongs to discover long-lived ultrapeers and in the future it will probably also use pongs to find GUESS (the new gnutella udp protocol) nodes and someday also CHORD (another udp based search algorithm) nodes.
 
Back
Top