FAQ: Re: key length

Bernhard Reiter bernhard at intevation.de
Fri Aug 8 09:45:03 CEST 2014


On Thursday 07 August 2014 at 17:16:26, Werner Koch wrote:
> On Thu,  7 Aug 2014 16:29, bernhard at intevation.de said:
> > You realize that your response is not an argument.
>
> Long time readers of this list understand what I am saying:
>
>   To evaluate security risks it is useless to pick one component of a
>   large system and believe you are done.
>
> and I also expected that you read today's security note.

I know and I did. Still your comment was close to cynism in the eye
of an unexperienced reader. ;)

> > I have given the datapoint to show that there are credible people
> > recommending 4096 bits. So it seem natuarl that less savy users
>
> and they are not talking about a system but an isolated algorithm.

So your argument is 
"Because the asymmetric crypto algorithm at RSA at 2048bit is usually 
not the weakest link in your system, we usually do not recommend more bits."?

So what is the common weakest link then? 
The symmetric cipher, the entropy source, the implementation issues in 
software and hardware (like side channel attacks)? 
My expectation is that we work together to get the default of a default
OpenPGP system to be so strong that is provides 10 years security.
At least my actions I can control.

> > ask us about it. Our argumentation must be better then this,
> > if we want to explain it to users and to convince them.
>
> Right, let's leave the bike shedding to others.

I am making an effort writing to this list because I want crypto to succeed, 
especially GnuPG and Free Software. The question about default key length
comes up so often, that a better explanation is helpful. So here I am 
exploring the arguments.


-- 
www.intevation.de/~bernhard (CEO)    www.fsfe.org (Founding GA Member)
Intevation GmbH, Osnabrück, Germany; Amtsgericht Osnabrück, HRB 18998
Owned and run by Frank Koormann, Bernhard Reiter, Dr. Jan-Oliver Wagner
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 473 bytes
Desc: This is a digitally signed message part.
URL: </pipermail/attachments/20140808/40542258/attachment.sig>


More information about the Gnupg-devel mailing list