Re: SERPENT implementation in debian
On Thu, Aug 26, 2010 at 11:52:31PM +0200, Christoph Anton Mitterer wrote:
> On Thu, 2010-08-26 at 17:43 -0400, Perry E. Metzger wrote:
> > I presume you mean the cipher by Anderson, Biham and Knudsen.
> Yep .... was there another one of the same name?
The linux kernel tree has crypto/serpent.c .
> > I'm unaware of one, but I think it is a fairly bad move to use
> > algorithms other than standard ones. Unless you have very good reason
> > to use something eperimental, I would stick to AES.
> I plan to use them stacked ... so that shouldn't be a problem IMO.
What do you hope to gain by that? Not performance. What do you hope to
get that you can't get from the standard AES-256?
> Any why should it be just experimental,... wasn't it very well analysed
> during the AES selection process?
Yes, it was. But the specific implementation you use wasn't as
well-analyzed as current implementations of AES are.
Tzafrir Cohen | email@example.com | VIM is
http://tzafrir.org.il | | a Mutt's
firstname.lastname@example.org | | best
email@example.com | | friend