[cap-talk] Dan Bernstein's qmail security lessons paper
Sandro Magi
smagi at higherlogics.com
Mon Dec 17 16:13:37 EST 2007
Jonathan S. Shapiro wrote:
>> Computers cannot represent true integer arithmetic because of the
>> infinity. ALL supposed computable "integer arithmetic" is modular
>> arithmetic [1].
>
> Nah. I'ld agree that all *correct* "integer arithmetic" is modular
> because of memory limitations, but I have yet to see a BigNum
> implementation in the wild that actually checks for this error. :-)
Well, most programs in the wild don't check for overflow either, they
just throw an exception or terminate; the similarities are striking! ;-)
Sandro
More information about the cap-talk
mailing list