Peter Thomassen <[email protected]> writes:

> We've been living this approach already, both slowly (with
> introduction of EC algorithms, close to what the ideal transition
> would have been) and quickly (SHA-1 phase-out).

There are no protocol police, so no bad decisions can be mitigated by
having a document.  I'm sure that's one thing we all do agree upon, and
there are many many many examples of intentional and accidental breaking
of various protocols over time.  Some by forced extensions that are
incompatible with the rest of the world for lock-in purposes, etc.

> I therefore have difficulty seeing what this document would change in
> practice; I'm also not convinced the document would have prevented
> RedHat from prematurely turning off SHA-1 support in their OS release,
> had it existed at the time.

It would be informational, to me, to have a document where we had the
discussion of what the right path steps and timing was, got IETF
agreement on it, so we could use it when making new documents like
RFC9905 (deprecating SHA1) and RFC9906 (deprecating GOST) so we would
have our own agreed up guidance to at least consider and hopefully
follow.

It will not prevent the need to jump multiple steps forward in
emergencies or other cases.  But it gives us a solid set of guidance to
at least try to follow.

Implementations and deployments will still force our hands at times.

-- 
Wes Hardaker
Google

_______________________________________________
DNSOP mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to