Re: [tcpm] Feedback on WG acceptance of draft-zimmermann-tcpm-cubic-01

["Scharf, Michael (Michael)" <[e-mail]>]

> On 2015-4-29, at 14:32, Scheffenegger, Richard <[e-mail]> wrote:
> > I'm still in support, on standards track.
> 
> Me too, given that this has been the default CC on Linux for years now.
> What's the reason for Informational; that this was done outside of the
> IETF and therefore doesn't deserve PS? That seems a bit dogmatic given
> the amount of real world experience.

In my view, there is sufficient real world experience, and we had quite a number of TCPM and ICCRG discussions on CUBIC in the past.

I actually care more about another aspect: The CUBIC algorithm have evolved in the past, at least regarding details, and it would be a surprise if this evolution would stop. 

I think the key question is whether we can come up with one "standardized" version of CUBIC that will hopefully be stable for some time, or whether the best we can do in TCPM is to document some snapshot of the algorithm. Both options seem possible to me in principle, but the former one requires more community effort to ensure that the document will not be outdated soon.

> But: I'd rather see it published as Informational that not at all.

In any case, there seems to be unanimous consensus that we will publish a document draft-ietf-tcpm-cubic.

Maybe a related follow-up question should be: What is the expected time-frame for the milestone of submitting draft-ietf-tcpm-cubic to the IESG? 2015? 2016? 2017?

Any thoughts?

Michael


PS: Bob raised a similar question on the AQM list: http://www.ietf.org/mail-archive/web/aqm/current/msg01159.html