> I don't blame you. But the fact is that not only is reseeding bad, > seeding at all is bad. The protocol already includes announcement and > initial discovery algorithms, and all the simulations have shown that > these give much better results then people randomly seeding. > > Ideally, all a node node should get is a single bootstrap reference.
ok, i experimented a bit with deleting the data store (and the routing table within) after only three nodes responded on the watchme network. i updated to the latest snapshot (1060900 bytes long) dated 2nd of june. i took the latest seednodes.ref (watchme). i configured the node for watchme mode. i ran the node. NONE OF THE 13 SERVERS RESPONDED ANYTHING. i did the same thing at my university, setting this node permanent and announcing. i added that reference to my seednode.ref, restarted my transient node and tried again. the public node accepted connections but rejected all requests. (i saw it in the logs) so, at the moment the watchme network seems not to satisfy even any single request. or do you make different observations? i only wonder why, because at first it worked fine and fast with the watchme nodes (very low node load too, compared with the normal network) - then two days later only three nodes responded and finally today nothing works. what the heck is going on? have i missed something? or are the 13 watchme nodes in the seednode.ref all down? mfg The Bishop _______________________________________________ devl mailing list [EMAIL PROTECTED] http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/devl
