Re: Re-revised idea for early cross-area review

[Andy Bierman <ietf at andybierman.com>]

Jari Arkko wrote:
> I've done a few early reviews, and do not think they are that
> expensive. Maybe even less expensive than full final
> reviews.
>
> It is true that the stuff is far from stable at -00. However,
> as a reviewer you need to have a different take on
> the review than what you normally do in the final
> stages. Namely, you should look at the architectural
> issues, missing pieces (this security scheme
> won't fly, you are ignoring congestion control),
> etc.

I would rather see the effort go into "Guidelines for Foo" RFCs
that we have been seeing lately.  Give the authors the info
they need to pass the review, without a time-consuming additional
review process.

I've never heard of forced review of a document that the WG
knows is full of problems, rough terminology, little or spotty WG consensus,
and will change significantly before it is ready for publication.
That's just about every draft-00 I've ever seen.  The cure for
the "waiting until IETF Last Call" syndrome shouldn't cost more
to solve than the original problem itself.

If a WG Chair or AD, who is doing a good job of overseeing a WG,
decides a particular draft needs an early review from another area,
then that's different -- that's just people doing their job.



Andy


> There's no need to look at the nits. Look at whether
> the idea is workable, and what other things the authors
> need to worry about when completing their work.
>
> I support Brian's review suggestion.
>
> --Jari