Hi...

the reason you might see fewer results is that other clients use differnt and advanced technics to route queries. Unfortunatly some of them are missing in Phex. But I'm trying integrate them sooner or later.

Limewire and Bearshare clients tend to cluster. This is because these clients use there own host caches and other clients have no access to these host caches. They tend to give out mainly IPs of other Limewires/Bearshares. Phex uses public host caches like Gnucleus is doing it. You are still not seeing a much smaller bit of the network. You only tend to have more none Bearshare/Limewire clients in your horizon.
Most of the Gnucleus clients you see are Morpheus clients.

The large send queue is do to a bug in the current Phex version. The next Phex version fixes this and is more stable in the network again.

The research behaivor improved very much since 0.6. It now reacts to different network and content availability situations. This dramaticly reduced the network pollution while still getting good results.

Gregor

__________________