Re: Bug fix release?

[Matthew Guenther <mguenthe engr UVic CA>]

On Tue, 6 Jun 2000, Pawel Salek wrote:
> 
> On 2000-06-06 10:59 Carlos Morgado wrote:
> > > I would /really/ like to work out this mailbox-update-detecting issue
> > > -- when the mailbox changes under Balsa, we do NOT handle it.  At all.
> > > I'm really hoping that there's no libmutt crapitude underlying this
> > > but I really want to work it out.
> > > 
> > 
> > i looked at that for like 2 nights, and my conclusion was "it should
> > work". then i noticed my conclusion was obviously wrong and i didn't
> > totally understand what was going on. 
> 
> Neither do I. I actually observe two issues: sometimes the deleted
> messages are not removed from the mailbox on the disk, and sometimes whena
> previously opened mailbox is marked as updated and I click on it, it turns
> to be empty, no even old messages. I have to close it, open it again
> (still no messages), close again, and open - and then it is finally
> alright. Do you guys understand that? (I admit I din't look much at this
> part of the code).
>

I've got somewhat of a handle on what's going on there, however I haven't
had time to code a fix lately.  One issue is that Balsa is closing the in
memory mailbox and then checking for unread messages on disk before the back
end has completed marking messages as read.  To demonstrate this I often
open the same mailbox in mutt and Balsa, then close the one in Balsa after
reading all the messages.  The mailbox will show up in the mailbox list as
containing unread messages, while in mutt you can clearly see they have all
been marked as read.  So this is simply a timing issue.  

As for updating the mailbox hash in memory when the file on disk is updated,
I have had a look at that and have a partial solution.  Again, I just need
to code something up, but haven't had much time lately.  I'll see what I can
do though.

As for the deleted messages not going anywhere and whole mailboxes
dissappearing, I haven't experienced either of those for awhile, I do
remember there was a problem with some experimental code I tried out that
caused things like this when the mailbox was synchronized to disk after
delete actions.  But I don't think that happens anymore.

MBG

-- 
Matthew Guenther                    If it doesn't work, force it. If
mguenthe@engr.uvic.ca               it breaks, it needed replacing anyway.
http://www.netcom.ca/~mguenthe/