http://blog.trendmicro.com/flaw-in-windows-vista-aslr-implementation/

lol dude take your midol

I didn’t even enter the job market yet.
Like I said, I do research at the
University and I’m very well paid.

Maybe you both quited from University? : \

PS: I don’t usually reply to morons,
specially when I’m under a fake alias and
have nothing to defend about myself.
But, you both need to see that nothing is
what it looks like.

Pinguin… Your wrong and thats why you got fired from your last job…

What selinux calls heap protection is a allocator algorithm. Basically a sort algorithm that works with frames. Simple shellcode payloads are defeated by it, but allocators have been around since the nineties and continue to be defeated through trampoline techniques and such. there have been a lot of xor allocators defeated through trampolining of native processes and tables both on linux and windows.

I guess technically it is protection, but I didn’t give that credit to windows either, windows just has better randomization and obscurity.

The software emulation of nx bit on both platforms are also allocators that just flag frames.

You might know something I don’t. I don’t monitor the innovations that much. I just know it’s not effecting my work and I have 9x and nt installs over and almost a decade old now. reformatting is like buying a new car when a battery needs replaced.

I’ve been a fan of openbsd for a long time and use it exclusively on servers and some laptops. I have about the same level of security on my nt machines as I do my thinkpad with an anally retentive configured openbsd 4.4 install on it. If people don’t have restrictions on processing and file systems the system is going to get hit hard even on top of perfect code and hardware.

surf

more

porn

OK, I’m sorry. This paragraph from you
made me understand that you made no idea
what you were talking about (linux):

“The stack protection in windows is way better compared to propolice(openbsd) and what’s in fedora 10. None of them protect again non-stack memory corruption like heap overflows etc.”

Since propolice is “only” a compiler flag
and rather rudimentary mechanism (it
only prevents a limited set of
memory corruption attacks),
I tried to explain what more is available
to linux systems. :)

So, my “you are wrong” statement was regarding
the fact that Linux has indeed (with
specific, non-vanilla patches) stack and heap
memory corruption protection that is
contradictory to your quoted claim.

I really don’t know much about microsoft dep,
but I can tell for sure that linux pax
can take advantage of the NX bit and emulate
it in case it is not present in the
hardware, which seems pretty much what
dep do.

“dep does better randomization and predictions along with pointer tracking.”
It is not my interest or either I have the
means to verify this claim, so I’ll believe
it for now.

“My overall point is that windows has equal or better security than linux, it’s just not enabled out of the box”
What is it useful for then, if it is
disabled? :P
That’s the main problem. Sysadmins *CANNOT*
be uninstructed people. IMHO Microsoft is
guilty on this one for their OS for dummies.

Believe it or not, the same is happening on
the opensource community. With the
introduction of ubuntu (which was not bad
at all), I’ve started to see the SNR level of project’s mailing lists decreasing a lot.

“I agree microsoft is a big scary corporation with greedy licenses and anti-trust issues, but we live in a materialistic dog eat dog world, and they’re just doing better at screwing other people over than everyone else through ease of implementation and practicality.”

I’m not with you on this one. While our
economy is capitalism driven, most of us
don’t care.
While money is good for living I belive we
should not live for money.
I do computer sciences research for a living
and I *hate* microsoft for what it is been
doing for all this years. it killed a lot
of good projects (either by buying them or
by lawsuit flood).
I entirely believe that we would be way
ahead in research and technology if microsoft
wasn’t here.
Considering the company size, it was expected
that they produce more innovation, instead
of copying/cloning everything that moves.

“Linux is nice and all, but try implementing it as a work station solution in a small-medium sized company and see how much it costs to manage and train around it. Most companies use it on backbones for this reason.”
Indeed.
I don’t think “linux is for everyone” (TM)
either.
I love linux because it oversimplifies
(as in methodology) software research.
I could tell you a lot of jokes about
colleagues of mine spending a lot of time
doing things on windows that would be done
in a matter of seconds using a few bash lines
on linux, like copy pasting unformatted data
to excel one-by-one at hand : \

It was nice to talk with you all,
thanks for your time.

I know about randomization and such, I read a credited article on shellcode attacks based on abstraction a long time ago, and it compared windows dep to stack defender, pro police, and selinux(platform irrelevant) etc..

dep does better randomization and predictions along with pointer tracking.

My overall point is that windows has equal or better security than linux, it’s just not enabled out of the box. You configure group policies and and enable dep a remote exploit has the same effect on windows as it does on openbsd 4.4 and fedora 10 default policies.

I agree microsoft is a big scary corporation with greedy licenses and anti-trust issues, but we live in a materialistic dog eat dog world, and they’re just doing better at screwing other people over than everyone else through ease of implementation and practicality.

Linux is nice and all, but try implementing it as a work station solution in a small-medium sized company and see how much it costs to manage and train around it. Most companies use it on backbones for this reason.

you are wrong.
fedora 10 comes with selinux which (if used
properly) applies aggressive MAC policies.
even if you could get an heap overflow,
you could not make it executable due to
selinux policies.

there is also grsecurity’s pax that prevents
stack and heap memory corruption using memory flagging. it also does heap and stack
randomization.
(latest vanillas also do stack randomization
for free).

please read more about it before starting
misleading people.