On Wed, 10 Oct 2007, Ken Murchison wrote:
If there are any outstanding issues that you believe still need to be
addressed in 2.3.10, please let me know.
Found the following this afternoon:
Note that the replication protocol currently does not have the facility
to support the IMAP CONDSTORE extension (modification sequences). It is
recommended that you do not try to use both CONDSTORE and replication
at this time. The deficiencies in the replication protocol will be
fixed in version 2.3.9.
I realise that we are perilously close to 2.3.10 now, but I believe that
http://www-uxsup.csx.cam.ac.uk/~dpc22/cyrus/patches/2.3cvs/modseq.patch
replicates modseq updates. It could do from review from someone who
actually understands CONDSTORE. I was in two minds whether to use separate
SetModseq and SetHighestModset commands, but I believe that all updates
bump the highestmodset, so I am using a single combined command to update
highestmodset and an arbitrary list of modseqs on the target mailbox.
It does involve a protocol change: an extra modseq item for every mailbox
and message from the USER and MAILBOXES command on the server, plus the
new SETMODSEQ command.
unexpunge doesn't update modseq: should it?
Also a trivial patch which should definitely be applied:
http://www-uxsup.csx.cam.ac.uk/~dpc22/cyrus/patches/2.3cvs/flags.patch
we are protected from an infinite loop by the:
while (msg && (record.uid > msg->uid))
msg = msg->next;
above in each case, but that's not really what it is there for.
--
David Carter Email: [EMAIL PROTECTED]
University Computing Service, Phone: (01223) 334502
New Museums Site, Pembroke Street, Fax: (01223) 334679
Cambridge UK. CB2 3QH.
