It's a combination of the two. Existing messages, that are already accounted for in courierimapuiddb, are obviously kept where they are. Anything that's left, any new messages, are then sorted by filename, before being added.
Ah... Okay, let me see if I've got this right then.
(First off, sorry for using "index" in the last message. I should have looked up my terminology first. -- Hope it didn't needlessly confuse anyone. I was talking about "message sequence numbers" not UID's, just to be clear.)
When Courier is building the list of message sequence numbers for a mailbox, it first reads in the list of "known" (i.e. not "recent") messages from the courierimapuiddb file. Then it sorts that list according to the UID's they have already been given -- rather than by their filenames or their order in the uiddb file.
Correct so far?
Then, it looks for "recent" messages (those in the Maildir, but not previously added to the uiddb file, right?) and independently sorts them by filename (which typically should have a number corresponding to when they were added to the Maildir -- making it effectively a date-based sort, assuming they have the expected naming convention, right?). Then it adds that new sorted list to the end of the previous sorted list of "known" messages.
Right?
Then it assigns to each message a message sequence number (the one it'll have at the beginning of the session anyway) based on its order in the full list.
Have I got it?
If so, then I retract my previous comments about unlinking/re-linking files to fix his "arrival" sort in PINE (which I think from his description is based on the message sequence numbers) ... And instead say if he wants to rearrange them for PINE's "arrival" listing, that he should -- disconnect all clients from that account of course, and then:
A) simply remove the courierimapuiddb file (*if* he can accept that his client will have to reload any and all cached messages), and
B) rename the files as suggested (in my message *before* the last one) with a numerical prefix based on a Received header (or Date header, if he so chooses) from each message.
That way, when he logs in the next time to the account, Courier will have an empty uiddb list, so that ALL messages in that mailbox will be sorted according to their filenames before getting their new UID's and message sequence numbers assigned.
Does that sound right, Sam? ... Thanks for helping to clear things up.
-jab
------------------------------------------------------- This SF.Net email sponsored by: Free pre-built ASP.NET sites including Data Reports, E-commerce, Portals, and Forums are available now. Download today and enter to win an XBOX or Visual Studio .NET. http://aspnet.click-url.com/go/psa00100003ave/direct;at.aspnet_072303_01/01 _______________________________________________ courier-users mailing list [EMAIL PROTECTED] Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
