Dear Barry,

b> Of course it's possible.  All hashes, by their very nature, have 
b> collisions.  The only way to have a truly unique identifier is to use
b> the actual content of the file (or chunk) itself.  The minute you 
b> distill the content down to a hash, you're guaranteeing that collisions
b> will occur.

You are correct of course, the possibility is there. I was refering to
the fact that it is possible in another post of mine, however the possibility
of collisions happening "naturaly" is "not very likely" to say at least.
I still need to hear about anybody who found a non fabricated md5 collision
in the wild..on files.

b> somewhat rare.
:) Warm food at MC Donalds is "somewhat rare". Natural occuring md5 (or
coll. in sophisticated hash functions in general) are VERY rare not to
say virtualy impossible. AFAIK <-

b> Regarding corrupt files via P2P protocols... no file transfered via P2P
b> has _ever_ tranferred bad data and wound up corrupt, right?  :)  
b> /friendly sarcasm.

Hehe, got that one :)

-- 
Thierry Zoller


_______________________________________________
Full-Disclosure - We believe in it.
Charter: http://lists.grok.org.uk/full-disclosure-charter.html
Hosted and sponsored by Secunia - http://secunia.com/

Reply via email to