Re: [TLS] draft-ietf-tls-oob-pubkey-06

[Nikos Mavrogiannopoulos <[e-mail]>]

On 11/03/2012 07:58 PM, Hannes Tschofenig wrote:


> I unfortunately missed the meeting last time so I am probably not the best person to reflect the discussion from the meeting myself.
> RFC 6091is an informational RFC and would require to be upgraded to a standards track document or a downref needs to be specified in the tls-oob-pubkey document. 
> Dealing with such a downref to RFC 6091 would be useful if RFC 6091 would actually be a good fit. There are two problems:
> 1. OpenPGP support in TLS is IMHO not deployed.  


Hello,
 I've never done a survey on the deployment, but I trust your research.

> 2. A problem I noticed with re-using RFC 6091 is that the certificate capability indication does not allow a client to distinguish whether it supports processing of OpenPGP certificates vs. whether it is able to provide OpenPGP certificates. For this reason the structure is different in draft-ietf-tls-oob-pubkey.
>>> I didn't see the need to add OpenPGP support in this document. 
>> So the idea is to have multiple certificate type extensions for TLS?
>> This was the meeting consensus?
> Luckily there are not that many certificate formats. If someone would want to define a new extension then they should be using draft-ietf-tls-oob-pubkey.


Shouldn't in that case the oob-pubkey draft obsolete or update rfc6091?
I don't think that both the certificate type registries should be active.

>> Why wasn't this brought up to the mailing list given that this overrides
>> the previous consensus [0] on the mailing list?
> I am not the chair of the group. I let them to judge consensus in the way they want. 


This was not a question for you.

regards,
Nikos