Re: edits to the draft should reflect the consensus of the WG

On Fri, Aug 17, 2007 at 11:24:22AM -0500, Dan Connolly wrote:
 
> It's our goal to reach consensus eventually, but I'm flexible
> on whether that happens before or after the editor updates the
> editor's draft.
> 
> In fact, the charter encourages the editor(s) to lead:
> 
> "We expect that typically, an editor makes an initial proposal, which is
> refined in discussion with Working Group members and other reviewers,
> and consensus emerges ..."
>  -- http://www.w3.org/2007/03/HTML-WG-charter.html#decisions
> 
> The process of working on issues is still rather chaotic; the chairs
> have not put *any* design questions to the WG formally, so don't
> have any established rhythm there. I don't have as good feel as
> I'd like for which issues the editor is putting priority on, and
> I can imagine lots of other WG members feel similarly and are
> frustrated about it.

I think there needs to be a process for tracking issues and placing
substantive proposals for changing the draft before the working group for
consideration and ultimately, determination of consensus.

A process by which the editors update the draft, and everyone else has to
search through the revised version (or a lengthy change log) to find out which
issues were and weren't addressed, and what changes were made, would make it
difficult to keep track of issues and formally decide on individual proposals
for change.

A better process would be to place proposals before the working group (for
example on the announcements list) after deliberation has taken place
surrounding the issue addressed by the proposal, and then call for objections
- proposals that are not objected to would be accepted by default, and a
  process would need to be agreed upon for dealing with proposals which
  attract dissent - for example, further discussion might be needed, a vote
  could be required if the issue remains contentious but there is a need to
  move forward, etc.

I am not supportive of any process that gives the editors unfetted discretion
to make whatever changes to the draft they think fit, based on their own
individual or collective judgments about the merits of the technical issues
involved. However, I would support a process that gives the editors limited
discretion to incorporate, on a provisional basis, changes that have been
discussed by the working group, which have been put forward as proposals, but
on which consensus has not been reached - these include proposals initiated by
the editors, the co-Chairs and other participants.

Sometimes, it is better to read a proposal in the context of a draft than to
read it as a separate mailing list post, which is why W3C working groups, in
my experience, sometimes request that the editors introduce changes into
drafts which are controversial, for the purpose of making it easier to
evaluate the impact of the proposals. I also think it is desirable for editors
to have discretion to make editorial changes that don't need to be
specifically placed before the working group for decision (before or after the
change is made), since these changes are generally stylistic and therefore
uncontroversial.

In summary, while I am flexible as to when changes are made to the CVS
repository, there has to be a process which makes it easy for everyone to know
what proposed changes are on the table, to decide whether to agree to those
changes, and to deal with proposals that don't receive consensus support - and
this process needs to be at the granularity of individual technical proposals,
not at the granularity of the entire working draft.

Received on Saturday, 18 August 2007 02:02:30 UTC