Re: The Standards Process

>Daniel Connolly:
>>...
>> No. Nobody pays for W3C specs. Members get them early, but nobody
>> pays for them.

This is too confusing, so I'll attempt to clarify it.  

W3C specs are not IETF specs unless the IETF chooses to make them so. 
IETF specs are not W3C specs unless the W3C chooses to make them so.
However, there are times when the features overlap, and in that case
the IETF specification generally comes *after* the W3C specification,
but only the IETF specification can define an Internet standard.

W3C members do not get W3C specs early, contrary to what Dan said.
W3C specifications are always published openly unless they contain
someone else's confidential information.

W3C members do have greater access to the editorial process, and to
the editors themselves, but that's mainly because we maintain an
open line of communication with W3C members, most of the people
who want to communicate with us are W3C members, and because W3C
members have signed an intellectual property agreement that prevents
them from claiming proprietary rights on new W3C ideas prior to
their publication in an open spec.

W3C draft pre-releases are viewable by members, and a selected few
non-members, before the general public can view them, but only when
there is a good reason for it (in most cases this is just to
verify that the changes made to the document are the same as those
requested by prior reviewers and the major stakeholders).  When
the document is published as a specification, it is available to anyone.

The W3C is trying to find ways to speed up the overall process.
To this end, W3C is toying with the idea of having a dedicated review
panel, sitting on a closed list, act as an editorial support group
for each topic area -- these people would commit the time necessary
for immediate feedback on ideas *prior* to specification changes.
In other words, they are giving immediate feedback on raw design issues.

As most of you will have noticed, I do not use a separate review process
for the HTTP 1.x specs.  The HTTP-WG gets the draft as soon as I have a
text version ready.  Sometimes I will give people pre-releases in the form
of a postscript file, but that is purely for my own convenience and is not
restricted to members.

The reason I don't use the new "W3C process" is because I already committed
to the IETF process for HTTP/1.0 and 1.1.  In many respects, you may
consider this to be a specification bake-off.  If I can get these
specifications out the door in the same timely manner as I did for
RFC 1808, then the WG will have proven that the goals of the new
"W3C process" (i.e., of being faster than the IETF WG process alone)
are not attributable to the unfocused discussions on the WG mailing list.

Naturally, it has yet to be seen whether this can be done.  Note also
that I will be returning to UC Irvine on September 29th.  This will
raise some thorny issues, since UCI is not a W3C member even though
I am a member of the W3C team.  It also means that I won't have any
grant support for my specification work.  OTOH, I have found there
to be many advantages to working as an independent, and I work a hell
of a lot more efficiently back in Irvine than I do in Boston.

If I cannot get the HTTP 1.x specs completed faster than the new
W3C process would complete a specification of the same quality, then
I suspect the IESG will shut down the current HTML and HTTP working
groups rather than have them stand in the way of faster progress,
and only reopen them for specific tasks.

What will all this mean for the future of WWW technology and the
history of openness in the WWW project?  I don't know.  However,
please do keep in mind that those of us being yelled at are the
same people that created the original open process.

 ....Roy T. Fielding  Department of ICS, University of California, Irvine USA
                      Visiting Scholar, MIT/LCS + World-Wide Web Consortium
                      (fielding@w3.org)                (fielding@ics.uci.edu)

Received on Monday, 18 September 1995 17:03:24 UTC