Re: [Dcrup] new version draft-ietf-dcrup-dkim-crypto-04

On Tuesday, August 08, 2017 02:46:22 PM Martin Thomson wrote:
> On 8 August 2017 at 01:51, John R Levine <johnl@taugh.com> wrote:
> > On Mon, 7 Aug 2017, Martin Thomson wrote:
> >> I don't believe that adding a signature algorithm means that you need
> >> to update 6376.
> > 
> > Well, OK, but where does all the new ABNF go?
> 
> In your draft?
> 
> It's not like you can't define new values, but it's also the case that
> you don't have to.  The already-defined extension point covers all
> valid values.
> 
> > Our plan here is that once this is published, people implement 6376 plus
> > this draft plus (or perhaps minus) Scott's anti-SHA1, which tells me that
> > these have to update 6376,
> 
> Yes, that's possible without updating 6376.  One problem I've
> encountered is that 6376 plus its updates and 6376 plus some other RFC
> are materially different to some people.  "Updates" is taken to mean
> corrections, and we're not really correcting it - it's perfectly fine.

I've never heard that interpretation.  I've seen updates used in the way we 
are discussing lots of times.  See RFC 7208 and RFC 7372 for a reasonably on 
the nose example.

Scott K

P.S.  Any document that considers 512 bit RSA keys OK, is not 'perfectly fine' 
in 2017.

> > but in any event, this is a decision for the WG
> > or the chairs, not for us.
> 
> I don't think that it's the chairs who decide these things, though I
> would certainly respect their views on this.  Ultimately, it's a
> community-consensus thing.
> 
> _______________________________________________
> Dcrup mailing list
> Dcrup@ietf.org
> https://www.ietf.org/mailman/listinfo/dcrup