-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Matthew Toseland wrote: > Hmmm, okay. Well will it - either the add-to-least-full-bucket or > add-to-most-full-non-full-bucket be fair if we only take into account > the number of tokens available, and not the queued/running requests?
I don't see why not - we're still controlling the number of requests each peer can send us via the token buckets. Cheers, Michael -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.3 (GNU/Linux) iD8DBQFEmwTryua14OQlJ3sRAm3pAKCoVhSmzEQjoCFLvX5tDLhOmLD/awCg7Xnk oKBgDKlyOKmdwRsWhRYnsF4= =J6bl -----END PGP SIGNATURE-----
