Re: Outlook HTML crash

From: Thor Larholm (thor@pivx.com)
Date: 03/17/03

  • Next message: Elkhatib, Ahmad: "Outlook Crashing, and not asking for password"
    From: "Thor Larholm" <thor@pivx.com>
    To: <http-equiv@malware.com>, <vuln-dev@securityfocus.com>
    Date: Mon, 17 Mar 2003 11:23:51 +0100
    
    

    This is definitely older than the latest IE patch, I've been experiencing it
    personally for more than a year. In my cases, I can reproduce the issue more
    easily when posting plaintext newsgroup message replies.

    Both HTML and plaintext messages are edited throught MSHTML, which is
    further exposed in IE (among others) as an ActiveX control (the DHTML
    Editing Component) and directly embedded through the
    designMode/contentEditable properties. Any editing that we manually perform
    in Outlook or Outlook Express which causes them to crash or overflow (like
    pressing enter in the replied-to text) can be programmatically reproduced.

    Regards
    Thor Larholm

    ----- Original Message -----
    From: "http-equiv@excite.com" <http-equiv@malware.com>
    To: <vuln-dev@securityfocus.com>
    Sent: Sunday, March 16, 2003 8:42 PM
    Subject: RE: Outlook HTML crash

    >
    >
    > That's very interesting. We were just discussing this last week, it
    > has been going on for months and seems to have come about after the
    > very latest Internet Explorer cumulative patch. Also, as Kurt points
    > out, hitting the enter key slowly or gently avoids the problem. Also
    > happens when you 'reply to sender is same format [html]' and manually
    > select plain text. Damn annoying "BUG!" in this particular vendor's
    > ever long quest to irritate people !
    >
    > MSIMN caused an invalid page fault in
    > module MSHTML.DLL at 015f:636bd8b8.
    > Registers:
    > EAX=004711a4 CS=015f EIP=636bd8b8 EFLGS=00010202
    > EBX=02d6f2ec SS=0167 ESP=0056f3bc EBP=0056f3dc
    > ECX=00000118 DS=0167 ESI=00000008 FS=124f
    > EDX=0040003c ES=0167 EDI=004711a4 GS=0000
    > Bytes at CS:EIP:
    > f3 a5 8b 4d 08 8b 49 0c 89 41 08 e9 76 13 fc ff
    > Stack dump:
    > 00000000 02d6fa90 0000001e 0355ab0c 00000000 00000000 00000000
    > 00000000 0056f474 6370f575 02d6f320 02d6fa90 0355ab0c 00000000
    > 00001e00 00000000
    >
    >
    > --
    > http://www.malware.com
    >
    >
    >
    >
    > they don't call it crapware for nothing.


  • Next message: Elkhatib, Ahmad: "Outlook Crashing, and not asking for password"

    Relevant Pages

    • RE: Calendar Time Zone, Outlook 2003 and KB928388
      ... Microsoft 928388 Daylight savings patch and Outlook 2003 appointments, ... they appear time shifted one hour, when viewed on a computer with Outlook ... 2003 with the patch and a computer with Office 2003 without the patch. ... To err on the side of caution, all clients connecting to the same mailbox ...
      (microsoft.public.outlook)
    • Re: Installing todays Office 2000 security update (KB892842) with
      ... Went back to updates now it only states that the Outlook Collaboration Data ... Objects (CDO) Update: ... Now if we can figure out how to get CDO update to stick (when I install it, ... patch) then it can include it in the file re-build. ...
      (microsoft.public.officeupdate)
    • RE: Calendar Time Zone, Outlook 2003 and KB928388
      ... impacting much more that a calendar. ... they appear time shifted one hour, when viewed on a computer with Outlook ... 2003 with the patch and a computer with Office 2003 without the patch. ... To err on the side of caution, all clients connecting to the same mailbox ...
      (microsoft.public.outlook)
    • KB835732 Errors and Fixes
      ... On our network we recently pushed this patch, ... administrator rights, the problems with Outlook, IE, etc ... The fix involves resetting the permissions under the WINNT ...
      (microsoft.public.win2000.windows_update)
    • Re: Ctrl+C doesnt interrupt process waiting for I/O
      ... some process is keeping the disk busy (some cron job for example: ... The following patch to 2.6.26-rc8 fixes the issue for me. ... I have tried to reproduce the original poster's issue on ... This disables the flush step, ...
      (Linux-Kernel)