[Imap-uw] ideas on improving efficiency of calls to mail_sort?

[Mark Crispin]

Hello -

The c-client mailing list has been renamed to the imap-uw mailing list.

The fundamental problem is that your server does not implement the IMAP 
SORT extension.  You would not have this problem with a server that offers 
SORT.

Next, what options are you giving to mail_sort()?  If you are setting 
SE_UID, then the code does not (can not) do any prefetching.  This is 
because c-client's cache is completely implemented by message sequence 
number, not by UID.  In general, options such as SE_UID and FT_UID should 
be avoided, especially when doing multiple message operations.

Instead, get the sequence numbers, and then afterwards translate them to 
UIDs (if that is what you want) via mail_uid().  This may sound backwards 
to someone who is thinking in terms of UIDs, but trust me, c-client cache 
will be happier and will reward you for your cooperation.

You may find the use of sequence numbers to be beneficial in other ways as 
well; use UIDs just for synchronization and do everything else with 
sequence numbers.  c-client will reward you for that as well.

-- Mark --

http://staff.washington.edu/mrc
Science does not emerge from voting, party politics, or public debate.
Si vis pacem, para bellum.