Re: Updated EXT_INFO draft - draft-ssh-ext-info-02

[nisse%lysator.liu.se@localhost (Niels Möller)]

denis bider <ietf-ssh3%denisbider.com@localhost> writes:

> (3) "Don't replace SERVICE_REQUEST + ACCEPT"
>
> I still believe SERVICE_REQUEST + ACCEPT serve no purpose, and waste a
> round-trip for no reason. 

I still fail to see why SERVICE_REQUEST costs a roundtrip time. IIRC,
you claimed or guessed that some implementations can't handle NEWKEYS +
SERVICE_REQUEST + UERAUTH_REQUEST sent back-to-back. But I don't think
I've seen any details. And I really don't think replacing
SERVICE_REQUEST is an appropriate or proportionate workaround for such
bugs.

> This seems like a great opportunity to do away with them.

It's fine to have a proposed change to replace SERVICE_REQUEST depend on
the proposed extension mechanism. But please don't introduce a reverse
dependency. 

I don't think there's likely to be consensus that we should do anything
about SERVICE_REQUEST now, so it's going to be an obstacle to the
extension mechanism which is a lot easier to agree on.

Regards,
/Niels

-- 
Niels Möller. PGP-encrypted email is preferred. Keyid C0B98E26.
Internet email is subject to wholesale government surveillance.