Re: Preventing Denial of Service Attack In IPC Serialization
On Jul 4, 3:53 am, jlind...@hotmail.com wrote:
7. Invokes operator () against rx, supplying the container and count
as arguments. << ***THIS IS THE PROBLEM***
I guess you didn't bother reading Sebastian Redl's post on the 26th of
June, or any of the followups.
I read them.
1. I pointed out that there was a flaw in my serialization framework,
and probably others.
2. Jeff pointed out that the flaw exits in Boost.
3. You attempted to refute that the problem exists in Boost.
4. I showed prima facie evidence that it does exist in Boost.
So, after so many posts, we at least arrive at a simple conclusion.
The problem does exist, and no dead horses are being beaten. Now, the
next step, is to find a solution, and not to kick at the one you
proposed, I have one gets more regular each day I think about it.
In any case, you should not claim that a problem does not exist when
it does. There are people who read these posts who do not understand/
care about all the details. They only want to know the end result:
does a problem exist or not. With your posts, it could have easily
been determined by a casual reader that there was no problem, and such
a reader might have used Boost Serialization in the nude, leaving
ample opportunity to crash their computers at will.
Do you at least admit that the problem exists in Boost (and MFC for
that matter)?
-Le Chaud Lapin-
--
[ See http://www.gotw.ca/resources/clcm.htm for info about ]
[ comp.lang.c++.moderated. First time posters: Do this! ]