W3C home > Mailing lists > Public > public-owl-wg@w3.org > October 2007

Re: document pubication schedule

From: Jeremy Carroll <jjc@hpl.hp.com>
Date: Tue, 23 Oct 2007 11:59:51 +0100
Message-ID: <471DD427.9090703@hpl.hp.com>
To: public-owl-wg@w3.org

** aside: I've just noticed the unfortunate typo in the subject line of 
this thread - my apologies for introducing it - too late to fix now.

Bijan said:
See this post from an experienced W3Cer for a similar view:

It does no harm, I believe, to have a clear starting point and has
the potential to do some good that is relatively difficult to get any
other way (i.e., wide early review).

I looked at Charles' message and agree with him about the value of the 
heart beat publication.

I don't believe anyone is arguing against that.

I think some of the concerns are:
- each potential reviewer is likely to review our documents at most once.
- if we ask for review of inappropriate documents, then we use up review 
effort that would be better spent on documents on which we have better 
agreement within the WG.
- the most appropriate way to indicate lack of consensus: many people 
seem to argue that publishing a document necessarily  indicates *some* 
degree of consensus about it.
- what sort of document will best get good review at this stage

So the three proposed docs are all very technical and hard for the more 
general reader to get to grips with. We all know that S&AS only got 
meaningful review from a handful of readers, and these documents seem to 
follow in the S&AS tradition of being written for the OWL specialist.
Most of the potential reviewers probably wrote the documents, and the 
other potential reviewers are in this Working Group, except for Dave 
Turner who has returned to Cambridge to write his PhD on some other topic.

Thus, for meaningful outside review, it is not a FPWD that is critical 
but a first publication of an accessible document that is critical. 
Whether this is a requirements doc, or a delta on the OWL Reference, or 
on the Guide, I don't know. I know it isn't the sort of document that I 
would read, let alone write - but then I am not the target reviewer.

I suspect that the sort of document Charles is referring to in his HTML 
WG message is something like an OWL Reference document.


A concern is that some members of the WG may expect WDs to have too high 
an initial quality. I think publish early, publish often, is a good 
mantra. As an example I point to:


where the SOTD says:
This is a first Public Working Draft of the Data Access SPARQL Query 
Language by the RDF Data Access Working Group (part of the Semantic Web 
Activity) for review by W3C Members and other interested parties. It 
reflects the best effort of the editors to reflect implementation 
experience and incorporate input from various members of the WG, but is 
not yet endorsed by the WG as a whole. Some sections are incomplete and 
there are a number of issues in the document and working group issues.

A (complete) example of an 'incomplete' section is

10 Summary of Query Patterns

Section status: Create this after preceding sections moderately stable.

Will be a brief summary of the terms defined above to bring them together.

Judging from the reviews so far, there is most consensus over the 
semantics doc, and perhaps that, alone, could be the FPWD.
I would advocate starting on a more accessible doc to be a follow on 
publication for say January, such as a delta on OWL Ref. Initially this 
can simply be new text, without any effort to integrate (see e.g.
[while personally I thought this poor practice for LC - presenting a 
diff - for a FPWD it seems an efficient use of time])

While Bijan has emphasized schedule pressure, I suggest that maintaining 
consensus is significantly more important. If only a few weeks into the 
WG, schedule pressure is bad, then I am afraid I simply say 'I told you 
so'. HP's position has always been that the schedule is unrealistic and 
conflicts with the need for consensus.

Seeing that there does appear to be consensus for some progress (e.g. 
publishing semantics doc reasonably soon) it would seem to me to 
unnecessarily stress the process to insist on publishing other docs 
before we had reached consensus on them.

Received on Tuesday, 23 October 2007 11:00:21 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:41:59 UTC