[Full-disclosure] Chinese Professor Cracks Fifth Data Security Algorithm (SHA-1)

[BlueBoar at thievco.com (Blue Boar)]

My understanding that the kind of birthday attack under discussion would
start at 80-bits if SHA-1 (at 160-bits) were 100% secure. The attack
under discussion is reported to reduce that to the neighborhood of
60-something bits.

I am not a mathematician though, so I would be perfectly willing to
believe I was wrong about that.

                                        BB

3APA3A wrote:
> Dear Blue Boar,
>
> It's  not  clear  if  this 'crack' cam be applied to birthday attack. My
> in-mind computations were: because birthday attack requires ~square root
> of N computations where bruteforce requires ~N/2, impact of 2000 times N
> decrease  for birthday is ~64 times faster. 64 = 2^6. Because complexity
> is ~square root of possible combinations, it's equivalent of traditional
> birthday  attack,  with  160-(2*6)=148  bits  hash (150 is my mistake in
> in-mind computations).
>
> Of  cause,  since  I  completely  wasted 10 years after obtaining Master
> degree  in  Mathematics  and  3 years after loosing last pencil I may be
> completely wrong in computations :)
>
> --Wednesday, March 21, 2007, 9:48:55 PM, you wrote to 3APA3A at SECURITY.NNOV.RU:
>
> BB> 3APA3A wrote:
> > > I  know  meaning  of  'hash  function'  term,  I  wrote  few articles on
> > > challenge-response   authentication   and   I  did  few  hash  functions
> > > implementations  for  hashtables  and  authentication  in FreeRADIUS and
> > > 3proxy.  Can  I  claim  my  right  for  sarcasm after calling ability to
> > > bruteforce 160-bit hash 2000 times faster 'a crack'?
>
> BB> Fair enough, your sarcasm tags didn't render properly in my MUA. I was
> BB> fooled by you stating that the birthday attack would be 150 bits.
>
> BB>                                           BB