Re: [Monotone-devel] Re: Policy branches - first steps

[Paul Crowley]

Jack Lloyd wrote:
> On Mon, Feb 26, 2007 at 01:17:24PM -0800, Justin Patrin wrote:
> > 3) DSA can only sign less than 256 bits (140?) so moving to SHA-256
> > and DSA will be problematic.
>
> DSA, algorithmically, can sign arbitrary size hashes, as long as you
> choose a large enough q parameter. Until recently the standardized max
> for q was 160 bits, FIPS 186-3 (currently in draft) allows longer q if
> you use a longer p (q=256 is OK with p=2048 or 3072 bits).

The question is how attached we are to ssh-agent.  It's very cool but 
it's limiting in this regard.  If we want to go with it, it's a question 
of whether we can persuade it to accept 256-bit q.  If we want to part 
from it then there are a number of signature standards we could 
contemplate, including ECDSA which is faster than DSA for the same 
security and has smaller keys.
--
  __
\/ o\ Paul Crowley, address@hidden
/\__/ http://www.ciphergoth.org/