7914 - [PP] [Tree widget onchange handler problem] Switching from one mailbox to another causes Messenger to quit

Status: VERIFIED FIXED

(SeaMonkey :: MailNews: Message Display, defect, P3)

Description

[Katsuhiko Momoi]

** Observed with 6/10/99 Win32 build **

I looked at some msgs in one mailbox (note: there are no highlights
with this build when you select a msg or a mail folder). After reading through
the msgs in the first mailbox, I tried to open a different mailbox.
That's when Messenger suddenly quits. This happens every time I
try the same thing.

Comment 1

[lchiang]

The highlighting problem is http://bugzilla.mozilla.org/show_bug.cgi?id=7917

Comment 2

[Katsuhiko Momoi]

I was using POP server setting when this problem occurred
and the 2 mailbox files are on my local hard disk.

Comment 3

[Katsuhiko Momoi]

Here's part of the TalkBack report I sent to Cyclone:

http://cyclone.mcom.com/reports/incidenttemplate.CFM?reportID=1099&style=0&tc=57
&cp=1&ck1=SUser+email+address&cd1=momoi%40netscape%2Ecom&ck2=SBug+ID&cd2=isNotNu
ll&bbid=9740191

There's an identical report earlier on this: Incident ID 9739872.

+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Trigger Type:  Program Crash

Trigger Reason:  Access violation
Call Stack:    (Signature = jsdom.dll + 0x6f15 (0x01086f15) ca587d9e)
jsdom.dll + 0x6f15 (0x01086f15)
jsdom.dll + 0x15559 (0x01095559)
js3250.dll + 0x1c08d (0x0032c08d)
js3250.dll + 0x16aa0 (0x00326aa0)
js3250.dll + 0x1337a (0x0032337a)
js3250.dll + 0x134aa (0x003234aa)
js3250.dll + 0x384a (0x0031384a)
jsdom.dll + 0x13c5c (0x01093c5c)
raptorhtml.dll + 0x840f0 (0x014540f0)
rdf.dll + 0x1bced (0x013abced)
raptorhtml.dll + 0x7d2d0 (0x0144d2d0)
raptorhtml.dll + 0x7d296 (0x0144d296)
raptorhtml.dll + 0x26394 (0x013f6394)
raptorhtml.dll + 0x86547 (0x01456547)
raptorhtml.dll + 0x26394 (0x013f6394)
raptorhtml.dll + 0x8666e (0x0145666e)
raptorhtml.dll + 0x6f3d5 (0x0143f3d5)
raptorhtml.dll + 0x6e6d7 (0x0143e6d7)
raptorhtml.dll + 0x26843 (0x013f6843)
raptorhtml.dll + 0x76a04 (0x01446a04)
raptorhtml.dll + 0x766b0 (0x014466b0)
raptorhtml.dll + 0x76690 (0x01446690)
raptorhtml.dll + 0x76621 (0x01446621)
raptorhtml.dll + 0x770f3 (0x014470f3)
raptorhtml.dll + 0x2fd75 (0x013ffd75)
raptorhtml.dll + 0x2d487 (0x013fd487)
raptorhtml.dll + 0x2cde7 (0x013fcde7)
raptorhtml.dll + 0x2cbe1 (0x013fcbe1)
raptorhtml.dll + 0x2c637 (0x013fc637)
raptorhtml.dll + 0x8bfc8 (0x0145bfc8)
raptorhtml.dll + 0x26843 (0x013f6843)
raptorhtml.dll + 0x1850b (0x013e850b)
raptorhtml.dll + 0x26843 (0x013f6843)
raptorhtml.dll + 0x17e49 (0x013e7e49)
raptorhtml.dll + 0x26843 (0x013f6843)
raptorhtml.dll + 0x8c8e9 (0x0145c8e9)
raptorhtml.dll + 0x29cf0 (0x013f9cf0)
raptorhtml.dll + 0x23611 (0x013f3611)
raptorhtml.dll + 0x22c27 (0x013f2c27)
rdf.dll + 0x1adfc (0x013aadfc)
rdf.dll + 0x1ef4f (0x013aef4f)
rdf.dll + 0x1568a (0x013a568a)
rdf.dll + 0x1a21c (0x013aa21c)
jsdom.dll + 0xf957 (0x0108f957)
js3250.dll + 0x1333d (0x0032333d)
js3250.dll + 0x16d08 (0x00326d08)
js3250.dll + 0x1337a (0x0032337a)
js3250.dll + 0x16d08 (0x00326d08)
js3250.dll + 0x1337a (0x0032337a)
js3250.dll + 0x16d08 (0x00326d08)
js3250.dll + 0x1337a (0x0032337a)
js3250.dll + 0x134aa (0x003234aa)
js3250.dll + 0x384a (0x0031384a)
jsdom.dll + 0x13c5c (0x01093c5c)
raptorhtml.dll + 0x84246 (0x01454246)
rdf.dll + 0x1bced (0x013abced)
rdf.dll + 0x1bd0f (0x013abd0f)
rdf.dll + 0x1bd0f (0x013abd0f)
rdf.dll + 0x1bd0f (0x013abd0f)
rdf.dll + 0x1bd0f (0x013abd0f)
rdf.dll + 0x1bd0f (0x013abd0f)
rdf.dll + 0x1bd0f (0x013abd0f)
rdf.dll + 0x1bd0f (0x013abd0f)
raptorhtml.dll + 0x84e98 (0x01454e98)
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Comment 4

[Phil Peterson]

Sounds like local mail; reassigning to putterman for M7.

Comment 5

[scottputterman]

reassigning to hyatt and cc'ing waterson.  This looks like a bunch of the
problems we've been seeing today.  It looks like most of the upper call stack is
in raptor and rdf.  Maybe the hyatt's changes that waterson backed out would fix
this.

Comment 6

[Katsuhiko Momoi]

** Checked with 6/10/99 Win32 re-spin **

No, the problem did not go away with the re-spin today. So,
it still crashes when I try to view msgs in another folder.

Comment 7

[Chris Waterson]

Hyatt's away for a couple of days. I'll take the bug for now.

Comment 8

[scottputterman]

As I said in bug 6459, my js code might be causing this to crash since I'm
depending on the node that got selected to be the event target.  If it's not
then that could be a problem, especially when switching to to a folder that has
no selection.  I'll fix the mail code when I switch back to using "onchange",
but I didn't think javascript was supposed to crash on accessing objects
incorrectly, but instead print out an error message.

Comment 9

[sol]

I haven't seen this problem, but if it still exists then I think it needs to be
fixed for PR1, so I added a note to the Status Whiteboard

Comment 10

[David Hyatt]

Fixed. Note that I have changed the event name from "onchange" to "onselect".

Comment 11

[esther]

Using 1999100320M10 on win98, 1999100118M10 on linux and 1999100120 on mac, I
don't crash switching and reading messages between folders of the same account,
or between accounts.  My system has 1 imap 1 pop and 2 mail and they all work.
I know this was a problem several weeks ago, since this was fixed a
month ago and I haven't run into the problem for a while I will verify this one.