W3C home > Mailing lists > Public > ietf-http-wg-old@w3.org > May to August 1997

Re: Revised charter milestones

From: Koen Holtman <koen@win.tue.nl>
Date: Wed, 28 May 1997 22:59:07 +0200 (MET DST)
Message-Id: <199705282059.WAA28780@wsooti08.win.tue.nl>
To: Larry Masinter <masinter@parc.xerox.com>
Cc: http-wg@cuckoo.hpl.hp.com
Larry Masinter:
>
>The area directors requested that we submit a revised
>charter (see below) by May 17. It's already late.
                        ^^^
Urgl.  I keep forgetting it is May already.  When I first read this, I
thought we still had >17 days.

>
>I think we have two clear milestones:

I would count updating 2109 to reflect implementation experience (the
state-man-mec draft) as another important milestone.  Actually, I
think our main milestones must be updating 2068 and 2109.  The rest is
secondary.

>
>- May 97 'Simple Hit Metering' to Proposed Standard
   ^^^

There are 4 days left in May, and we cannot complete a WG last call in
4 days, so this has to be June at least.

Two pieces of information are still missing with respect to hit
metering:

 - Roy was opposed in the latest last call, I don't think we have seen
   him comment on the latest draft yet.
 - We have no clear view of how many implementers want to implement
   this thing.

I think we are not ready to go to last call before these pieces of
information are in place.

>- September 97: HTTP/1.1 (revisions of 2068 & 2069)
>  to Draft Standard
>
>The 'editing group' notes were rather terse. The question
>is: are there any realistic dates 

The September date for revising 1.1 depends for a large part on how
much time the `editing group' can dump into editing 1.1 in this
period.  To determine if September is realistic, I would need to hear
some statements from the members of this group about their personal
schedules.

>for any other documents
>that are not part of HTTP/1.1?
>
>"all non-passed milestones have dates...in the future, and 
>that the group thinks are realistic."
>
>So, can we schedule PEP, transparent content negotiation,
>revising state management...etc. realisticly?

See above on state management.

As for PEP, I have no idea, but I would say 3 months as a lower
bound.  I am not that happy with the latest draft, and I am a bit
concerned about all radical revisions between drafts.

As for TCN (the main draft), I have stated before that, due to
implementer pressure, my goal as an editor is to freeze this thing
within N months.  I am currently polling implementers to determine
this N.

What needs to be determined after this poll is in what form TCN could
realistically be frozen, given the N month timescale.  There are 3
options: Proposed, Experimental, or Not A Product Of This WG.

I think we can make Proposed, actually.  Though a lot of people in
this WG have no opinion on TCN, or have the opinion that they won't
implement it but don't mind if other people do, I have not seen anyone
claim yet that TCN is actively bad.

Of course TCN does not solve all negotiation problems, but this is not
a requirement.  Hit metering does not solve all demographics problems
either.

So far for my ideas on the milestones.  This sample size of 2 shows
that there are some diverging views around, but I hope we will be able
to converge on a milestone list soon.  What we need now is more input
on this topic.

Koen.
Received on Wednesday, 28 May 1997 14:01:14 EDT

This archive was generated by hypermail pre-2.1.9 : Wednesday, 24 September 2003 06:32:43 EDT