summaryrefslogtreecommitdiff
path: root/mail/vm
diff options
context:
space:
mode:
authorghen <ghen@pkgsrc.org>2006-08-18 06:46:05 +0000
committerghen <ghen@pkgsrc.org>2006-08-18 06:46:05 +0000
commit97c774a78ecf86ad906f54f572a1283464da50f0 (patch)
tree0336389f66c633bdd23533710ff75b651beec787 /mail/vm
parentcb301f27951d41fcc3bfaf91c36fdf17f1ad87de (diff)
downloadpkgsrc-97c774a78ecf86ad906f54f572a1283464da50f0.tar.gz
Update to dovecot-1.orc7. From the release announcement mail:
> Can everyone now agree that there are no more hangs? :) > > * Require that Dovecot master process's version number matches the > child process's, unless version_ignore=yes. Usually it's an > accidental installation problem if the version numbers don't match. > * Maildir: Create maildirfolder file when creating new maildirs. > > + ldap+prefetch: Use global uid/gid settings if LDAP query doesn't > return them > + %variables: Negative offsets count from the end of the string. > Patch by Johannes Berg. > - kqueue ioloop code rewrite > - notify=kqueue might have caused connection hangs sometimes > - deliver: If message body contained a valid mbox From_ line, it > and the rest of the message was skipped. > - mbox: We got into infinite loops if trying to open a 2 byte sized > file as mbox. > - Don't crash with ssl_disable=yes > - quota plugin caused compiling problems with some OSes > - mbox: After saving a mail to a synced mbox, we lost the sync which > caused worse performance > > I think my v1.0 TODO list is: > > - avoid duplicate flag change notifications, or in case the message is > also expunged don't bother notifying its flag changes at all > (shouldn't be hard) > - HFS+ avoid-hardlinks patch? > - 32bit -> 64bit upgrade still doesn't work without assert-crashing: > mail-cache-transaction.c: line 709 (mail_cache_add): assertion failed: > (fixed_size == (unsigned int)-1 || fixed_size == data_size) > - Courier-compatible INBOX. namespace gives "invalid namespace" errors > > I'm still not sure about the last one though. Could someone give me > specific commands that clients send that causes it? > > As for the 32bit -> 64bit upgrade fix, could someone give me access to a > x86-64 machine for a while to test it out?
Diffstat (limited to 'mail/vm')
0 files changed, 0 insertions, 0 deletions