I did a few more changes and it seems to be working with FreeBSD. I didn't before realize that event filters weren't bitmasks so that broke things. Anyway, a bit more "real world testing" would be better than my test program and telnet localhost -tests :) I'll release 1.0rc7 in any case today before I go to sleep (so, max. 6 hours or so). http://dovecot.org/nightly/dovecot-latest.tar.gz -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 191 bytes Desc: This is a digitally signed message part URL: <http://dovecot.org/pipermail/dovecot/attachments/20060817/1b7ee951/attachment.bin>
On Thu, 2006-08-17 at 21:06 +0300, Timo Sirainen wrote:> I did a few more changes and it seems to be working with FreeBSD. I > didn't before realize that event filters weren't bitmasks so that broke > things. > > Anyway, a bit more "real world testing" would be better than my test > program and telnet localhost -tests :)Actually, forget about this for a while. Also because they weren't bitmasks now I understand why the original code actually worked. Anyway, I'm still doing a bit of cleanups. Will build a new snapshot soon. -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 191 bytes Desc: This is a digitally signed message part URL: <http://dovecot.org/pipermail/dovecot/attachments/20060817/ae0e53ef/attachment.bin>
Well, I already built it before you sent the second message. Everything seems to be working normal so far...nothing abnormal in the logs and the only processes are from people who are actually connected :-) yayay rc7 will rock Timo Sirainen wrote:> On Thu, 2006-08-17 at 21:06 +0300, Timo Sirainen wrote: > >> I did a few more changes and it seems to be working with FreeBSD. I >> didn't before realize that event filters weren't bitmasks so that broke >> things. >> >> Anyway, a bit more "real world testing" would be better than my test >> program and telnet localhost -tests :) >> > > Actually, forget about this for a while. Also because they weren't > bitmasks now I understand why the original code actually worked. Anyway, > I'm still doing a bit of cleanups. Will build a new snapshot soon. > >