Openwall GNU/*/Linux - a small security-enhanced Linux distro for servers
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Fri, 16 Mar 2018 13:32:29 -0400
From: Matt Weir <cweir@...edu>
To: john-users@...ts.openwall.com
Subject: Re: Is john really using Narayanan and Shmatikov whitepaper?

As a note of caution, I'm 99% sure that Simon's presentation at
Passwords12 was about advancements to JtR's Markov model that could be
made vs. what's actually implemented, though the sections about
nbparts, (slides 7-14), should apply.

A good reference that I had mentioned to Matlink privately, but I
figure I should post here for everyone else, is the JtR Wiki write-up
on Markov mode: http://openwall.info/wiki/john/markov

Side note, I had completely forgot about Simon's work to provide an
accurate index function for JtR Markov that he detailed in his
presentation which would be useful to the previous thread Matlink had
brought up :)

Alexander, talking about incremental, are you aware of any write-up
for it behind it besides "look at the code?" Something like the JtR
Wiki article for Incremental might help other researchers. In practice
I've found Incremental is much more effective then Markov so I tend to
steer people to it as well, but there's not much out there to
understand what's going on under the hood. I know I've struggled with
that, (and made mistakes describing it).


Matt

On Fri, Mar 16, 2018 at 10:35 AM, Solar Designer <solar@...nwall.com> wrote:
> Hi,
>
> On Fri, Mar 16, 2018 at 02:23:04PM +0100, Matlink wrote:
>> I was wondering, for research purposes, how John was applying Markov
>> chains for its markov mode.
>
> I think the closest to documentation of Markov mode's internals is this
> presentation, by Simon who contributed this mode to jumbo:
>
> http://www.openwall.com/presentations/Passwords12-Probabilistic-Models/
>
> I also want to say once again that if you do any research involving
> JtR jumbo's Markov mode, please also include in your research JtR's
> incremental mode, which pre-dates Markov mode and isn't named Markov but
> is also closely related.
>
> Unfortunately, the Markov vs. incremental mode comparison in the above
> presentation uses JtR 1.7's incremental mode, whereas 1.8's should
> perform better.
>
> Alexander

Powered by blists - more mailing lists

Your e-mail address:

Powered by Openwall GNU/*/Linux - Powered by OpenVZ