From: Marc G. Fournier (no email)
Date: Wed May 01 2002 - 07:20:39 EDT
Is there any chance that the 'spam' extension to sieve is going to get
added to the CVS? *cross fingers*
On Tue, 30 Apr 2002, Lawrence Greenfield wrote:
> Hi,
>
> I'm happy to announce that Cyrus IMAP 2.1.4 has been released. This
> has several useful enhancements and bugfixes, and all users of 2.1 are
> encouraged to upgrade. This is the version we're using internally at
> CMU (modulo last minute documentation) and I encourage people running
> 2.0 or older to consider upgrading.
>
> I've included the changes below.
>
> Download Cyrus IMAP at
>
> ftp://ftp.andrew.cmu.edu/pub/cyrus-mail/cyrus-imap-2.1.4.tar.gz
>
> Complaints, comments, suggestions, and especially bugfixes are welcome
> on this list or to .
>
> thanks for all the help out there (and especially you Ken!),
> Larry
>
> ---cut here---
> Changes to the Cyrus IMAP Server since 2.1.3
>
> * All "MAIL" and "SIEVE" notifications are now handled by notifyd
> which is a daemon that supports multiple notification methods. The
> mailnotifier and sievenotifier options have been added to
> /etc/imapd.conf to configure notifications. (Ken Murchison)
> * Many feature enhancements and bugfixes for the Cyrus Murder. The
> code now supports live (but not transparent) moving of mailboxes
> from one server to another.
> * Some warning fixes.
> * fdatasync() is no longer required.
> * Fixed a bug in imap/append.c that would show itself if a message
> was being delivered to five or more different partitions.
> * Deliveries now don't create a redudant temporary file using
> tmpfile(); the staging directory is used instead. (Ken Murchison)
> * Fix a possible crashing bug in squatter. (Ken Murchison)
> * Deleting a user now also removes their Sieve scripts.
> * cyrusdb_skiplist: release locks during iteration. Should prevent
> denial of service attacks and possibly increase performance.
> * cyrusdb_skiplist: introduce a new mode using O_DSYNC writes which
> is possibly faster on Solaris. Currently off (it seems to hurt
> performance on Linux).
> * master has preliminary code to avoid forking storms.
> * sieveshell should now loop through all available SASL mechanisms
> before conceding defeat.
> * sieveshell can now upload a file to a different name.
>
>
|
|
|