Interesting time line:
2009/04/07 Identification of the vulnerability.
2009/04/23 Notification of IBM via the customer.
2009/04/23 Technical knowhow exchange between scip AG/IBM.
2009/06/05 Asking for current status by scip AG. (no answer)
2009/07/09 Asking for current status by scip AG.
2009/07/09 Reply with current status and assigned PMR.
2009/08/24 Asking for current status by scip AG. (no answer)
2009/09/08 Public disclosure of the advisory.
Published by IBM today. If you need to fix, you have to contact IBM Support.
this might be a stupid question but if you have to log a call wont you need a support contract ? or does that not apply to security fixes
Another stupid question - HTML formatted emails are rendered by Internet Exploder too.
That must also be done in the local context. If you have ever printed an HTML email from Notes 8+ you will know that you get the standard IE print header and footer complete with the local URL in the footer.
Someone please tell me there isn't a code execution possibility right there.
Best bet is to open a call and find out. (I am not sure if you can or not). Either way there are 3 other options you can do in the meantime.
By the way. The Technote was not published today (only updated today). It was posted before scip AG posted the exploit publicly.
@Chris - Actually, HTML formatted emails as "cleaned up" first to prevent malicious code. The cleanup may not be perfect, but it is a different scenario.
@Ben is that clean up exactly documented somewhere?
I fear you are both missing the point. The problem is that the Notes RSS reader, dumps the HTML into a local file on the user's PC. When IE is then invoked to render that local file, it does so outside its sandbox.
When IE, as a browser, renders HTML received from an HTTP source, it does so in its sandbox. Now one could argue about how effective that sandbox is, but that is a whole other issue!
@Nick: Neither Chris nor Ben talked about HTML from an HTTP source, but from HTML included in formatted mails from outside. I think it's not too unlikely the same happens with them.
But BTW: isn't that what every other local RSS reader application would do?
No, they don't. It's a very basic design failure. Other feed readers like FeedDemon use the proper Internet sandbox.
If you start digging around in the "standard" client outside of what used to be Notes, you will find lots of stuff that will lead to more security discussions. Code signing, ECL etc don't exist there. And you are supposed to turn malware protection off in this space, so that performance does not suffer.
Well, who's to bless & who's to blame ? One of the biggest questions on opening the Notes Client to the Eclipse platform for me was about security of plugins. The monolitic C++ Notes Client was much less vulnarable to lazy programmers than the open Eclipse platform is. That IBM itself had to prove that is some sort irony to me. Eclipse is a blessing and a curse - flexibilty and openness always means responsibility... . And - don't get me wrong - even with this, I appreciate the way, IBM is heading with Notes 8.x, Eclipse and Expeditor.
It's probably a matter of trust. If you want to use widgets, you have to trust them. If you can't trust a plugin from IBM, whom do you want to trust? The RSS widget is a perfect malware dropper. Access to the local file system, ability to run processes in user space.
@Vowe: So when I use a feedreader with offline capabilities (I think most of the popular do), I would load the HTML from the local file system as well?
Lotus Notes RSS reader is a joke anyway.
Volker Weber on Zu schwierig für Google at 00:40
Lutz Lengemann on Und Ihr wundert Euch at 22:23
Tobias Hauser on Zu schwierig für Google at 21:57
Dirk Stelloh on Zu schwierig für Google at 18:46
Peter de Haas on CIO describes how he moved 125K workers to Office 365 in 6 months at 16:56
Volker Weber on You may want to close this barn door on your IBM Connections site at 11:41
Oliver Regelmann on You may want to close this barn door on your IBM Connections site at 10:15
Matthias Röder on What you don't read on this site at 08:54
Volker Weber on How Sonos is getting better, year over year at 08:33
Felix Binsack on How Sonos is getting better, year over year at 08:04
John Keys on How Sonos is getting better, year over year at 06:08
Michael Sampson on What you don't read on this site at 02:25
Ray Bilyk on What you don't read on this site at 23:48
Paul mooney on What you don't read on this site at 21:33
Joe Litton on What you don't read on this site at 21:27
Frank Mueller on New Apple keyboard and trackpad at 21:04
Frank Mueller on New Apple keyboard and trackpad at 21:02
Theo Heselmans on What you don't read on this site at 20:41
Pascal Decker on What you don't read on this site at 20:34
Volker Weber on How Sonos is getting better, year over year at 18:14
Ingo Seifert on What you don't read on this site at 17:53
Volker Weber on What you don't read on this site at 17:41
Theo Heselmans on What you don't read on this site at 17:28
Alan Lepofsky on What is the future of Notes/Domino? at 15:29
Volker Weber on How Sonos is getting better, year over year at 15:18