Re: MD5 Alive?
- From: JuiceMan <jaysgeneral@xxxxxxxx>
- Date: Wed, 22 Aug 2007 19:36:55 -0700
On Aug 22, 7:53 am, "Sebastian G." <se...@xxxxxxxxx> wrote:
Ertugrul Soeylemez wrote:
MD5does not, and in fact cannot, guarantee uniqueness. If, like stated
above,MD5with the same input space as output space is a bijection,
then it does guarantee uniqueness for inputs of exactly 128 bits, but we
don't know that even for the originalMD5.
Actually we should assume quite the contrary: IfMD5is a pseudorandom
function, than it's a pseudorandom mapping and therefore about 1/e of all
outputs will not occur, and about the same number will be collisions.
Thanks for the feedback. I figured monkeying around with the internals
of MD5 would not help me.
However, some alarming points were brought up and I wouldn't mind
clarification on them.
It sounds like I got a big NO to my 25 times, split, 25 more times. As
in this is NOT making MD5 any more secure and in fact might be making
it less secure.
Running the Digest on itself say upwards of 1000 times is -- in the
circles that I've been in kind of an accepted thing to do.
For example if I take "abc" and run MD5 on it, take the answer run MD5
on that, and do so a 1000 more times what I get is no more secure then
if I ran it once?
If I'm reading correctly it even sounds that what I get utlimately at
the end of 1000 times might be the same as if it had been done with
"xyz" 1000 times?
Is this true?
- Prev by Date: Re: MD5 Alive?
- Next by Date: Re: MD5 Alive?
- Previous by thread: Re: MD5 Alive?
- Next by thread: Re: MD5 Alive?