Are WDs still needed? (was: Re: Meet&Greet with HTML and CSS WGs)

* Karl Dubost wrote:
>What is the notion of mature? Do you mean the publication of the Last 
>Call, or the fact that they will publish a Last Call and might go again 
>for a second Last Call?

My impression is that the Last Call maturity level nowdays means that
the Working Group believes the document is ready for review by other
Working Groups and other interested parties, while the Process document
suggests that it means the document is ready for Recommendation status
only subject to changes that might be necessary due to implementation
and testing feedback in the Candidate Recommendation phase.

This is quite obvious if you look at http://www.w3.org/TR/ and check
for each recent document whether there was another Working Draft for
the document after the first Last Call announcement. For example,

  charmod         -> 3 Last Calls
  webarch         -> 2 Last Calls
  rdf-concepts    -> 2 Last Calls
  xptr-framework  -> 2 Last Calls
  qaframe-spec    -> 2 Last Calls
  css3-text       -> 3 Last Calls *
  css21           -> 3 Last Calls *
  css3-selectors  -> 2 Last Calls *
  ccxml           -> 2 Last Calls *
  svg12           -> 2 Last Calls *
  svgmobile12     -> 2 Last Calls *
  xpath-functions -> 2 Last Calls *
  xslt20          -> 2 Last Calls *
  ...

Where * means that I expect at least x Last Calls even though there were
only x - 1 Last Calls so far. This development is most unfortunate, most
Working Groups do not really plan for multiple Last Calls of a document,
so having multiple typically (significantly) delays the specification
which reduces trust in the Working Group's ability to deliver their work
timely.

This also wastes time and causes frustration of reviewers, as you know,
multiple recent Last Call Working Drafts did not even meet basic quality
standards as put forth in W3C Manual of Style, the Character Model, the
QA Specification Guidelines, etc., people who rely on the maturity of
the document (book authors, other standards organizations, implementers)
are not unlikely to later need to change their work due to changes that
were introduced late in the process, etc.

This becomes ever more a vicious circle as people anticipate the first
Last Call for a document won't be the final Last Call anyway and so they
plan to review the document late, and for normal Working Drafts they do
not send comments as they typically have reason to believe that those
comments will be ignored anyway. Working Groups consequently, as you
point out, begin to plan for multiple Last Calls...

At some point it might indeed become pointless for a Working Group to
publish Working Drafts before the Last Call Announcement, and in fact,
e.g. SMIL 2.1 started its life on the Rec track with Last Call...
-- 
Björn Höhrmann · mailto:bjoern@hoehrmann.de · http://bjoern.hoehrmann.de
Weinh. Str. 22 · Telefon: +49(0)621/4309674 · http://www.bjoernsworld.de
68309 Mannheim · PGP Pub. KeyID: 0xA4357E78 · http://www.websitedev.de/

Received on Wednesday, 9 March 2005 21:57:02 UTC