Re: Conclusion of the last call on draft-housley-two-maturity-levels

[Mykyta Yevstifeyev <[e-mail]>]

10.09.2011 16:56, Barry Leiba wrote:
>> 1) Did the IESG consider processing this as RFC 3933 process experiment?
> How on Earth could that possibly work?

Those who are interested in something working will certainly find the 
way to make it work.

>
> First, simply the fact of the experiment will almost certainly prompt
> people to participate, resulting in a number of specs upgrading from
> PS to IS during the experiment... regardless of whether that pattern
> would continue afterward.  The experiment would be entirely tainted by
> its own existence, and would appear to succeed regardless of whether
> the change will actually help in the long term or not.
>
> Second, should we decide that the experiment failed and we do not want
> to continue the process change, what happens to all the IS documents
> that advanced from PS during the experiment?  Are they rolled back to
> DS?  Do they stay at IS?  In the former case, what does that do to
> assumptions that had been made on the basis of IS status?  If the
> latter case, how is that fair to documents that didn't have the
> opportunity to "skip" a rung in the ladder?  And wouldn't that be even
> *more* of an incentive for people to push their docs from PS to IS
> during the experiment, exacerbating the first effect?

So, for the purpose of the experiment, there will be a "2nd level", 
which is Draft Standard now and would be Internet Standard.  People 
wanting to participate in the experiment will be warned that a "2nd 
level" of Internet Standard might be of temporary nature and therefore, 
in the case of experiment failure, will be Draft Standard.  
Correspondingly, RFCs advanced to "2nd level" of IS, in the case of 
permanent adaptation, will be Internet Standard, and in case of 
returning to 2026 model, will be Draft Standard.  The term "experiment" 
implies that something will go in some way which is unusual; this is 
that unusual way.

Anyway, do we need to care what maturity level has some spec reached?  
Eg., YAM has undertaken an effort to advance RFC 4409, mail submission 
protocol spec, to Full Standard (now 4409bis is approved as FS), but, 
I'm sure, this won't affect the many implementations which use and will 
continue to use mail submission protocol independent on its current 
maturity level.  SMTP spec is now on Draft Standard, and I don't think 
it is now more widespread that when its spec was on Proposed Standard.  
Does that matter?

>
> Doing this as a process experiment doesn't make any sense; we either
> make this change, or we don't.  And we need to stop wasting time
> arguing about it.

"We either make change, or we don't" - unless there is a stable 
consensus on the change, the second variant is going to be the right one.

BTW, stopping arguing is indeed a good idea; starting discussing is 
indeed a better idea (if we haven't yet).  So what we're doing - 
discussing or arguing?

Mykyta Yevstifeyev

>
> Barry
>