Karl Pearson wrote: > > On Wed, March 17, 2010 1:55 pm, Tomas Kuliavas wrote: >> >> >> Karl Pearson wrote: >>> >>> I've emailed a couple times, with the most recent just a week or so ago >>> about Squirrelmail not reporting the correct number of emails left >>> unread >>> in all my folders. >>> >>> I solved it by un-checking this option under Folders: >>> >>> Enable Cumulative Unread Message Notification: >>> >>> I guess I don't know what that means, so clicked it years ago when I >>> first >>> started using SQM. Anyone want to fill my brain up with knowledge? It's >>> pretty much empty right now, which is obvious to most. >>> >> >> I though that you talked about invalid count reported by newmail plugin. >> Left listing counts are different. >> >> Do you have some folders that have similar names? Could you check which >> folders report invalid unread/total counts or make a spreadsheet with >> folder >> names, unread/total counts reported by IMAP STATUS and unread/total >> counts >> reported on left folder listing? >> >> Cumulative count should increase unread/total email counters, if folder >> has >> collapsed subfolders and those subfolders have unread emails. I suspect >> that >> counter can be incorrectly increased, if folder has collapsed subfolders >> and >> there are similar folders. > > Unchecking the Cumluative setting didn't solve it after all, so I'm seeing > the count creep up after just a couple days. > > I've been using IMAP-based email solutions on my own email server for over > 15 years, and a cursory check (ls -R|wc -l) shows I have 8540 separate > folders (lines; some aren't folders). I've found 15 duplicate names, so am > going through and either combining them, or renaming them, whichever is > appropriate. > > If renaming some of the folders (none of the duplicates were in the same > location) doesn't solve it, I'm wondering if a buffer overrun might be the > problem with that many folders. > > I wrote a script some time ago that goes through a folder (from the > commandline) and marks all messages as having been opened (Read). I do > that on older folders because I suspected at the time of writing it that > Alpine and Squirrelmail were 'fighting' over how to actually mark a > message as having been read. From what I'm learning, that's a function of > Dovecot, right? The email client sends an IMAP command and Dovecot marks > the message as opened/read, right? > > But I still have the question as why using the Squirrelmail 'Search' > feature would cause the unread count to be reset to 0 (even when there are > some unread emails). > > I'm running the MarkEmailRead.sh script against some folders and it takes > some time, so I think I'll head to bed and let it finish... > Which counters are increasing? You have lots of folders, but you still notice increased counts. SquirrelMail has mini plugin, which can show only folders with unread emails. Do you use some server side or SquirrelMail filters? SquirrelMail search resets Recent count. SquirrelMail does not issue any special command that asks server to change flag. It only opens mailbox in RW mode. This action should automatically reset Recent count in IMAP. Seen flag can be modified only when SquirrelMail changes flag (STORE id:id +FLAGS (\Seen)) or when code reads email body (FETCH id BODY[]) When folder names are similar, incorrect counts can be displayed only for folders that have collapsed subfolders. In any other case SquirrelMail does basic "STATUS mailbox (MESSAGES UNSEEN RECENT)". I am not 100% sure about it and I will have to do some tests in order to confirm it. If you mess with mail store directly, do you invalidate dovecot cache when you change something? If Alpine messes with your mbox files, set it to use IMAP. -- Tomas -- View this message in context: http://old.nabble.com/Message-Count-Update-%28Solution-Bug%29-tp27936550p27943036.html Sent from the squirrelmail-users mailing list archive at Nabble.com. ------------------------------------------------------------------------------ Download Intel® Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev ----- squirrelmail-users mailing list Posting guidelines: http://squirrelmail.org/postingguidelines List address: squirrelmail-users@xxxxxxxxxxxxxxxxxxxxx List archives: http://news.gmane.org/gmane.mail.squirrelmail.user List info (subscribe/unsubscribe/change options): https://lists.sourceforge.net/lists/listinfo/squirrelmail-users
