Openwall GNU/*/Linux - a small security-enhanced Linux distro for servers
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Fri, 10 Aug 2012 14:09:24 -0500
From: "jfoug" <jfoug@....net>
To: <john-dev@...ts.openwall.com>
Subject: RE: SRP

This protocol is trivially explained in
http://www.reddit.com/r/gaming/comments/xypw2/blizzard_security_update/c5qx7
70

Piece of cake to implement.  The bottleneck will be in the GMP expmod
speeds.  But with N being only 256 bits, the expmod is very fast (but much
slower than SHA1).  I used to do a lot of work with GMP (and some other fast
exponentators), when I was coding on OpenPFGW.  For 256 bit size, and N
being a 'non-special' number, GMP is about as fast it gets, for general
purpose.

Any chance of finding the leaked s, u, v data, to know if a format would be
working properly?

Jim.

>From: Solar Designer [mailto:solar@...nwall.com]
>
>Hi,
>
>Here's our opportunity to start supporting SRP verifiers:
>
>http://www.reddit.com/r/gaming/comments/xypw2/blizzard_security_update/c5qx
770
>
>Anyone?

Powered by blists - more mailing lists

Your e-mail address:

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