[Full-disclosure] Steve Gibson smokes crack?
bkfsec at sdf.lonestar.org
Fri Jan 13 21:57:36 GMT 2006
Jason Coombs wrote:
> The Microsoft corporate entity may not be malicious in terms of
> purposefully planting backdoors with knowledge and consent of Gates et
> al (this assertion is of course questionable) however, individual
> programmers at Microsoft have probably planted backdoors on purpose.
> This happens frequently in many software shops.
Oh I'm quite certain that it happens...
> The corporate culture at Microsoft made it easy to do so, and get away
> with it, as you so accurately described. Individual product managers
> who encouraged the least safe configurations and least safe
> feature/code designs might have done so for the purpose of preserving
> widespread access to such backdoors.
Perhaps... it's really tough to tell the difference. My assertion would
be that it can be difficult to tell the difference between an accidental
bug, a design flaw, and an intentionally planted bug. Of course, that
would depend on the bug and any evidence in the code regarding the bug,
but unless there's something that says "My exploit here", as sort of
happened with the NSA backdoor fiasco, it still might be difficult to
prove. Even then, we still don't know that that was an NSA backdoor
beyond a shadow of a doubt. There are worms out there with copyright
notices listing the government of China. Did China actually create the
worm? Why would it put a copyright notice in the code? More likely
that data is there for the purpose of deception. So even comments and
symbols aren't 100% trustworthy. (Not the same scenario, but still
illustrates that trust is difficult)
I think we need to be careful about making accusations without solid
I know that you don't like the concept of prosecution without solid
> It would be relatively simple for Microsoft to determine whether any
> particular individuals were responsible for writing the bad code and
> deploying flawed architectures over and over again through the years.
Assuming they made more than a handful of blatant and patterned holes
that had been found. There are ways to circumvent the infrastructure
such that they're not obvious. Inserting extra code into an old code
tree parser, for instance, so that the native code is trojaned but the
source isn't. That's only the first method that comes to mind.
It would depend on the intelligence and planning of the individual.
> Perhaps Microsoft has bothered to look into this by now, and has
> quietly dismissed the perpetrators.
Again, hard to say. Without being personally privy to any information
saying as much from a source who can claim to have been there to witness
it, I can't even speculate.
> Beware of ex-Microsoft programmers.
I would say that this is true for any ex-programmer of any widely used
program. Even if they didn't maliciously insert a backdoor, it's still
entirely possible for that person to know and understand the weaknesses
in the infrastructure. All infrastructures have weaknesses somewhere.
Full-Disclosure is hosted and sponsored by Secunia.