Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Date: Tue, 7 Dec 2010 08:50:52 +0300
From: Solar Designer <>
Subject: Re: pwgen

On Thu, Dec 02, 2010 at 12:24:11PM -0600, Minga Minga wrote:
> The 'random' passwords for the DEFCON contest were generated by pwgen.
> But I have long since lost the command line.

My guess is that you used the "-s" option:

       -s, --secure
              Generate completely random, hard-to-memorize  passwords.   These
              should  only be used for machine passwords, since otherwise it's
              almost guaranteed that users will simply write the password on a
              piece of paper taped to the monitor...

Yesterday, I generated a .chr file from 1 million of "pwgen -s"
passwords, and I started an attack on random-1000-from-pwgen.txt
(NTLM-hashed) using that .chr file.  It cracked 4 passwords so far:

0:04:21:37 + Cracked u0
0:10:42:05 + Cracked u561
0:14:18:03 + Cracked u223
0:17:24:04 + Cracked u151

01j1eL0Z         (u0)
Wi28bpuE         (u561)
9YjnhqjN         (u223)
6R5d5Pr5         (u151)

guesses: 4  time: 0:19:31:08  c/s: 16196M  trying: DNc8ErG6 - DNc8Err9

Considering the time it'd take to search the entire keyspace at this
speed, this means that those passwords are in fact just as secure as
they can be given the character set and length, at least against this

Thus, if my guess re: your use of "pwgen -s" is correct, then passwords
generated in this way are safe (although passwords of this type are not
safe enough when processed with a very fast hash, as it can be seen
above).  It's only pwgen's "pronounceable" passwords that are much
weaker than they look.


Powered by blists - more mailing lists

Confused about mailing lists and their use? Read about mailing lists on Wikipedia and check out these guidelines on proper formatting of your messages.