Message-ID: <200403261805.i2QI5O7p043273@grenada.globat.com>
From: mvp at joeware.net (joe)
Subject: Microsoft Coding / National Security Risk

 
Most likely there was testing and research to make sure no one was using the
function internally in a way that wasn't intended that could possibly break.
Of course I am completely guessing. It could have been so that MS could take
their time and hack into all of the open source groups using Windows
machines... 

Back to the serious world, I have several key modules that I use in many
different of the tools/programs I write. I am more careful updating them
than any others because they are so central to everything and all of my code
is written and leverage internally by just me. Since my code goes back 10+
years I can't guarantee that I *always* did everything exactly the same and
correctly, so I make an effort to take a quick peek at what I did do any
time I touch central modules. 

I think that much code would be hard for anyone to manage. Especially when
MS didn't actually write all of it, they go out to vendors for code as well
as write things internally plus there is stuff that they purchased that they
slowly convert over (previous examples would be like Zoomit or Sybase SQL). 

Again I don't like the fact that it took so long to patch, I do notice that
MS is working harder and harder to be secure over the last several years
though so I will give them the benefit of the doubt that they are working
towards making that better.

  joe
 
 

-----Original Message-----
From: full-disclosure-admin@...ts.netsys.com
[mailto:full-disclosure-admin@...ts.netsys.com] On Behalf Of Frank Knobbe
Sent: Wednesday, March 24, 2004 12:07 PM
To: joe
Cc: full-disclosure@...ts.netsys.com
Subject: RE: [Full-Disclosure] Microsoft Coding / National Security Risk

On Wed, 2004-03-24 at 07:06, joe wrote:
> [...] They weren't correcting a single self-contained program like 
> W3SVC or Apache or netdom, they were correcting functionality in a 
> core component used widely across the OS.
                  ^^^^^^^^^^^^^^^^^^

But it's just that -- A core component. Not hundreds of core components.  It
was ONE DLL that needed fixin', not a multitude of them.

I think the Windows source code has grown to a size that is hard even for
Microsoft to manage. I'm not surprised if the very developers are starting
to lose trust in their own code because it has grown to galactic
proportions... hence the need to extra long test cycles.

Regards,
Frank




> 
> If the next slammer virus came through and started formatting hard 
> drives, I would say the same thing I did when the last one came 
> through and that would be "How come you weren't patched with a patch that
had been out that long?".
> It doesn't matter how fast MS produces patches if admins and users 
> aren't getting them applied. The issue isn't simply one of technology, 
> it is also one of process. A vast number of people don't want 
> automatic updates of their systems either because they don't trust the
source or simply don't'
> want their machines updating automatically but DON'T go back to do it 
> in a conrolled fashion. They wait until someone says they need to go 
> do it. I don't let MS update my PC automatically but I do make it a 
> point to go back and check every couple of days to see if something 
> has been released and I watch several notification streams as well. 
> Most people will not do this so they either need to go with some form of
automatic updates or unplug.
> 
> MS sent many many people through the code. People outside are going 
> through the code. Again this isn't one program that one person could 
> go through and have a strong handle of. I don't think 10 more people 
> could add much if any value. Not sure 100 outside people could. If 
> this were the case we wouldn't be finding old holes in other open 
> source now, we would only be finding stuff in the newly released code. 
> I would however like to think that MS is working on better automated 
> scans of the code to find holes, that would be more value than trying 
> to find 10 excellent security programmers. I am someone who has access 
> to the current source and have walked through large sections of it, it
isn't like the holes jump out and say "HI, here I am".
> Also the code I have had a chance to walk through in the last 8 months 
> is pretty decent, I definitely am not going, oh my god oh my god. It 
> seems more rigorous than the code I have walked through say for Samba 
> however that is an objective opinion and am not going to enumerate 
> items I think one does better than the other.
> 
> BTW, how many zero day exploit based worms/viruses have been beating 
> up on MS in the last year or two... Not being flip here.
> 
> 
>   joe
> 
> 
>  
> 
> -----Original Message-----
> From: full-disclosure-admin@...ts.netsys.com
> [mailto:full-disclosure-admin@...ts.netsys.com] On Behalf Of Richard 
> Hatch
> Sent: Wednesday, March 24, 2004 5:10 AM
> To: full-disclosure@...ts.netsys.com
> Subject: [Full-Disclosure] Microsoft Coding / National Security Risk
> 
> Hi all,
> 
> Microsoft have stated that to make the source code for Windows 
> publically available would be a risk to National Security.
> Microsoft also took 9 months to produce a fix for the ASN.1 problem.
> 
> As much as some people may regret it, Western civilisation runs on 
> Microsoft software.  Imagine the panic that would ensue if the next 
> slammer worm infected 10 machines then formatted hard drives, or 
> scrambled random parts of random files.
> This is not news, some old DOS viruses set file lengths to zero, 
> rather than deleting files that could be recovered.
> 
> So my idea is this:
> Take a team of really really good C/C++ coders with excellent security 
> vulnerability knowledge and have them go through the source code for 
> windows (starting with the core functionality and internet facing 
> functionality maybe).  Find these bugs (including methodical black-box 
> testing against the
> binaries) and fix them.
> 
> These people would be fully supported by Microsoft (including full 
> access to all technical documentation, Microsoft technical advisors, 
> etc), and backed by the NSA or other Government agency.  Microsoft 
> could impose whatever NDA's they want, but they should fund the bug hunt.
> Not only can they afford it, they created the problem code.  Fresh 
> insight into how Windows functions is required to identify the less 
> obvious vulnerabilities.
> 
> Microsoft Windows is not just another piece of software, it has become 
> a fundamental part of businesses and governments.
> 
> Oh, can anyone suggest a reason why disclosing the source to Windows 
> would be a National Security risk, yet Microsoft is happy to provide 
> the same source code to ceratin third-parties (I assume this means any 
> company that has enough cash and signs the right paperwork).
> 
> Folks, simply reacting to 0days just doesn't work.
> 
> R. Hatch
> 
> 
> 
> ---
> 'The mirrors have grown vast and beautiful and very very *hungry*' 
> 
> The views and comments expressed in this email are the personal views 
> and opinions of the author and should in no way be considered an 
> official statement/release of QinetiQ.
> 
> Neither the author or QinetiQ can be held liable for actions taken 
> based on the information contained within this email.
> 
> _______________________________________________
> Full-Disclosure - We believe in it.
> Charter: http://lists.netsys.com/full-disclosure-charter.html
> 
> _______________________________________________
> Full-Disclosure - We believe in it.
> Charter: http://lists.netsys.com/full-disclosure-charter.html

Powered by blists - more mailing lists