W3C home > Mailing lists > Public > w3c-rdfcore-wg@w3.org > June 2003

Re: Cut back RDFCore semantics doc

From: Brian McBride <bwm@hplb.hpl.hp.com>
Date: 18 Jun 2003 11:38:09 +0100
To: Jeremy Carroll <jjc@hplb.hpl.hp.com>
Cc: rdf core <w3c-rdfcore-wg@w3.org>
Message-Id: <1055932689.2677.38.camel@dhcp-91-136.hpl.hp.com>

I have been having some discussions offlist, seeking guidance on how
best to proceed.  Briefly to provide some context:

I am becoming increasingly concerned about the health of the WG.

  - mail traffic has been down significantly this month
  - we have lost 3 telecons in about the last 6-8 weeks
  - we failed to get a quorum last Friday
  - I'm worried that folks aren't committed enough to get finished
  - as long as we are here unfinished, we are getting in the way of new
work getting started and progressing the SW vision.
  - we are about to head into the summer holiday season where progress
inevitably slows.

Looking at what needed to be done, I was particularly concerned about
the semantics doc and the fact that we are still getting bug reports on
it.  Lately those bugs have centred on an informative section of the
document, and I've been thinking about the most effective way to get the
normative parts to rec.

One idea was move the informative and contentious part (section 4) to a
note so we can the rest of the document moving.  Another approach would
be to weaken the claims made for the closure rules, i.e. that they are
complete.  I personally am relatively comfortable with either approach.

Publishing as a note has the following advantages:

  - we can take a little more time and check that the completeness claim
is correct, without holding up the REC track (only matters if we believe
completeness is important)

  - I think a note is easier to update than a rec.

Publishing a weakened claim for the closure rules has the advantage that
they are in the spec immediately.  I understand that some folks, e.g.
HP  would prefer the closure rules remain in the spec, are not concerned
about completeness but would rather we got done soon than risked the WG
running out of steam.  The disadvantage is that updating a rec is more
work, and the existence of rules in a rec discourages separate
publication as a note.  I think a note is more flexible.

On Tue, 2003-06-17 at 20:06, Jeremy Carroll wrote:
> > removing the closure rules? I'm totally unconvinced
> > and feeling completely lost... (again tears)
> > better later publish an errata!

I'm sorry Jos that you saw this without the context.

> 
> I have discussed this issue with HP colleagues and we support Jos's position.

Interestingly, I too have discussed this issue with HP folks and I got
nowhere near so strong a reaction as "totally unconvinced" or
"completely lost".

> 
> While we obviously should not rule out cutting the closure rules, we should be 
> convinced that it is necessary to do so 

That is not what your colleague said to me.

> 
> Given that they are informative I would argue that even if we are not 100% 
> convinced that they are right it is OK to go to Rec acknowleding the 
> possibility of errata.

My concern is that we can't move forward with known errors.  The problem
is not that we might have errors - all specs might have errors.  The
problem is that we are out of time, and we need a way to stop the
continuous feed of errors we are receiving.

I believe Pat is working on two versions of the semantics document.

  a) one has the closure rules excised
  b) the other has the closure rules moved to an appendix with the
    claim of completeness removed.

Both of these approaches address my concerns.  I am immensely grateful
to Pat for responding so quickly and comprehensively.

I would like quickly to choose one of these options.

My old project manager instincts are to cut right back to the
essentials, but I hear HP and Jos (and I think Pat) prefering option b,
which I consider riskier but acceptable. Does anyone wish to argue
against that?

Brian
Received on Wednesday, 18 June 2003 06:38:29 EDT

This archive was generated by hypermail pre-2.1.9 : Wednesday, 3 September 2003 09:57:56 EDT