Killing old/slow groups - transition thinking

[John C Klensin]

Dave,

--On Friday, 13 December, 2002 14:11 -0800 Dave Crocker
<dhc@dcrocker.net> wrote:

> John,
> 
> Wednesday, December 11, 2002, 1:42:00 PM, you wrote:
> John>   suppose we adopted a lighter weight review
> John> process for PS.  In doing so, we made it very clear to
> the John> community (and vendors, etc.) that, while we would
> _document_ John> known loose ends and exercise reasonable care
> to avoid putting John> things out that would cause serious
> damage, we rather expected John> PS documents to have bugs,
> 
> Let's take the considerations you are pursuing, and look for
> choices that do not try to reverse an existing effect.  The
> industry deals with PS RFCs a certain way, and it seems very
> unlikely that we can change it.  We cannot make anything "very
> clear to the community" that is different from what they now
> think.

I guess I'm a little more optimistic than you are.  Having been
on the industry decision-making side of this several times,
there  are industry entities out there (certainly not all of
them) who would prefer to wait for full standards if those
actually appeared, and appeared in reasonable time.  Well, that
isn't going to happen.  But I believe that, were we to shift the
PS norms and WG behaviors so that industry typically saw 

	* PSs within 12 months after work initiation, but with
	clear indications that they were likely to change with a
	little more thought and experience, and
	
	* DSs within another 6 to 12 months with both solid
	writing and solid implementation experience behind them,

we would see some behavior shifts.  Your guesses and mileage may
differ, of course, but I'd note that this gives "industry" (and
everyone else) DS documents sooner than we are typically getting
out PSs today (hence the proposals from you and others about 18
month rules).  Having been in several of the relevant
discussions, one thing that causes people to ship product at PS,
or even with only I-Ds, is the perception that Proposed takes
forever and that Draft takes several times forever.  Change the
timing, and change the perceived polish/guarantee level, and I'd
predict that some industry behavior would change.

Of course, this would do nothing to help the organizations that
are driven by their marketing organizations to issue press
releases that strongly imply that informational documents are
full standards and that posting an I-D implies IETF endorsement
of their technologies.  Nothing will fix that (other than,
perhaps, legal action by defrauded customers), but that syndrome
should not prevent us from looking at real issues and
possibilities.

> The consideration you have raised is that try to deal with
> proposals that are clearly or possibly not ready for prime
> time. That is, we do not have rock solid comfort in their
> technical viability. So we want to float them in the industry
> and see how they work, or what changes they need.
> 
> Hmmm.  Sounds like Experimental to me.

I think there are many cases where we should be using
Experimental and have been using PS instead.  In my opinion, the
IESG has not been taking a hard enough line in that area, i.e.,
that "market pressures" for something with a "standard" label on
it have been too high. 

I also note that the description of PS in 2026 does not require
"rock solid comfort in their technical viability".  It requires
only "generally stable, [has] resolved known design choices,
[is] believed to be well-understood, [has received] significant
community review, and appears to enjoy enough community interest
to be considered valuable".  And there are not supposed to be
any "known omissions" unless the IESG examines those and decides
to release it anyway.  I.e., there needs to be a general belief
that it is probably ok, but "rock solid" is not a requirements.

     john