[plug] john the ripper

[Matt Kemner]

Hi Bill

> Thanks Matt, all good info, most of which has been not obvious in the
> sources I checked.  One discrepancy - you say you can do 100,000 c/s
> with john but I am showing much less - can you confirm the figures you
> give?

Yes, but that's using old-school DES encryption for the passwords, not
this new-fangled MD5 thing :)

As I said it was a lifetime (or two) ago that I last used john on a
regular basis.

live:~/JOHN# ls -la john
-rwxr-xr-x    1 root     root       137980 Jul 13  1999 john*

live:~/JOHN# ./john -incremental oldshadow
Loaded 31 passwords with 31 different salts (Standard DES [48/64 4K])
guesses: 0  time: 0:00:00:05  c/s: 65868  trying: jwk - comessam
guesses: 0  time: 0:00:00:11  c/s: 90263  trying: shadial - cattema
guesses: 0  time: 0:00:01:42  c/s: 106695  trying: zhi - 9oi

And from ./john -test :
Benchmarking: Standard DES [48/64 4K]... DONE
Many salts:     111692 c/s real, 112821 c/s virtual
Only one salt:  106112 c/s real, 106112 c/s virtual
 ...
Benchmarking: FreeBSD MD5 [32/32]... DONE
Raw:    1958 c/s real, 1954 c/s virtual

> running without nice gives a solid 2916 c/s each time

Interesting that you only get 50% better speed than this old PIII

> Found some docs installed with it and it appears that it doest use all
> possible characters by default - seems like something simple has turned
> "difficult"

Trying all characters should be (or at least was) the default, but you can
always force it with:

john -incremental:all

>From doc/EXAMPLES:

"In the configuration file
 supplied with John these parameters are to use the full 95 character set,
 and to try all possible password lengths, from 0 to 8."

 - Matt(who should probably go and grab a more recent version for purposes
        of this discussion ;)