http://bugzilla.spamassassin.org/show_bug.cgi?id=3225
------- Additional Comments From [EMAIL PROTECTED] 2004-04-23 02:51 ------- Subject: Re: RFE: Bayes optimizations -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 > > we did find that the precision of the atime values was valuable for > > effective expiry. > > I must be missing something. How is the time more than a rough guide to > deciding which tokens to expire? Do tokens accumulate so quickly that > expiration has to be performed more often than every two hours? What > didn't work when you tried the two-hour granularity? We found that it adversely effected expiry, causing it to fail to be able to expire due to too many messages being combined into the same "bucket", adversely effecting expiry's ability to find a good cut-off. Now that I think of it, though, we may have been using the *scan* time instead of the message's origination time (from Received hdrs); this may not be an issue with the latter... The best thing to do here is to test it out, using mass-check I think. - --j. -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.4 (GNU/Linux) Comment: Exmh CVS iD8DBQFAiCxDQTcbUG5Y7woRAv9lAJ9GZinMMEIfF+Uv/s76AFe19kSXwgCfURU9 M7V6PPkLRzgwl9Ka4bbukWM= =gydk -----END PGP SIGNATURE----- ------- You are receiving this mail because: ------- You are the assignee for the bug, or are watching the assignee.
