On May 2, 2013, at 07:21, "Eggert, Lars" <lars(_at_)netapp(_dot_)com> wrote:
Yeah, all kinds of issues, but if we created a new thing here in between WGLC
and PS, the broader industry would never understand.
That is a matter of naming and marketing ("candidate RFC"?).
The "this is baked, go and implement it" signal to the outside is only one
aspect, though.
The other two things that were mentioned:
-- get the cross-area review going;
-- get the numbers (IANA).
As we are seeing e.g., with MPL (ROLL WG) right now, the latter can go wrong
(aspects of the protocol still change, the allocation has to change with it).
But that isn't too much of a cost in most cases.
Re the cross-area review: Doing it earlier doesn't mean doing more. In theory.
In practice, not-really-baked stuff comes up and needs to be reviewed n times.
(As an appsdir reviewer, I have seen that happen.)
I think the WGLC is the wrong process step to define something as baked.
-- too late, because there may still be some (identifiable) random corner that
the WG wants to fix;
-- too early, because the WG may be deluded or mismanaged.
Much better would be a simple collusion between WG chairs and sponsoring AD.
(In any sensible WG, you would still involve the WG, of course.)
Grüße, Carsten