127695 - failure when writing two documents to IMAP sent folder simultaneously

Status: RESOLVED WORKSFORME

Description

[Dan Wallach]

I'm using the UW imapd in the default 'mbox' mode, so it's slow when those 
files get big.  If I compose a message and send it, it takes as much as ten 
seconds to actually write the document to the 'Sent' folder.  No sweat, because 
everything else still works.

Except if you try sending a second message before the first one has completed.  
It will successfully deliver the mail message to the SMTP server, but it will 
generate a dialog box indicating it failed to talk to the IMAP server and 
giving you the chance to go back to the composition window.  What should really 
happen, I suppose, is to either open another connection to the IMAP server or 
to queue the IMAP request behind the one that's pending.

A secondary but related issue: when this happens, if you say "yes", take me 
back to the compose window, and then try to kill the window, it gives the 
normal warning that the message has not been sent, BUT IT HAS!  You need to 
keep some state around for this.

Comment 1

[Dan Wallach]

FYI, I'm running Mozilla 0.9.8 on Win2000.

Comment 2

[Jim Laurent]

This is compounded by bug 129495 particularly over dial-up lines with large Sent
mail folders.  The long delay in writing to the sent folder begs the user to
sent more mail while waiting.  Result is that there is NO way to get the failed
message into the sent folder without sending it again.

Comment 3

[Max Kanat-Alexander]

[e-mail] --

Does this still happen on a recently nightly build? It's been quite a while
since this bug was filed. :-)

-M

Comment 4

[Dan Wallach]

I haven't seen the bug lately, but we recently got a much faster machine for our
IMAP server.  On the other hand, now Mozilla Mail is randomly wedging or
crashing far more often, with no real discernable pattern.

Comment 5

[Max Kanat-Alexander]

All right. Well, do you think that you could close this bug, but try to work out
what's causing your MozMail to crash now, and file a new bug with that? If you
file it well, CC me on it and I'll see if I can confirm it.

-M

Comment 6

[Dan Wallach]

It's up to whoever the bug is assigned to (Scott MacGregor) to decide whether to
close it or whatnot.

As to my most recent mail bug, if I can only figure out how to reproduce it,
I'll file something.  I've managed to generate several TalkBack things, but I
don't know how those ultimately filter down to filed bugs.

Comment 7

[Max Kanat-Alexander]

All right. :-) Actually, since the bug is unconfirmed, you can close it if you'd
like, since you're the reporter. I'm pretty sure that (especially with Netscape
engineers) the engineers rarely look at Unconfirmed bugs, even if they are
assigned to them.

-M

Comment 8

[Max Kanat-Alexander]

Oh, and do you know what IMAP server is on the other end? Sometimes Mozilla has
problems with certain IMAP Servers, and it'd be nice to narrow down what
problems are related to what. :-)

-M

Comment 9

[Jason M. Abels]

Actually, it's not assigned to somebody until it's assigned. New or Unconfirmed
bugs can be marked as WFM by the submitter.

Going ahead and marking WFM.