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

Working Drafts, Recommendations, Working Group Notes etc.

From: Jeremy Carroll <jjc@hpl.hp.com>
Date: Thu, 29 Nov 2007 16:43:22 +0000
Message-ID: <474EEC2A.3010104@hpl.hp.com>
To: "Web Ontology Language ((OWL)) Working Group WG" <public-owl-wg@w3.org>


There is occassional discussion of whether some document or other should 
be Rec Track or not ...

I thought it might be worth stating what I feel are general principles 
that I have seen used in the passed.

The normative definition of the language must be in rec track documents, 
which end up as Proposed Recommendations, and then hopefully 
Recommendations.

(A working group can define things in WG Notes, but these are not 
'recommendations' but merely work in progress).

For other material (i.e. 'informative' material) there are a range of 
choices for the final state of the material. One way to proceed is to 
put such material into documents called Working Drafts.
Having published a working draft a working group eventually does one of 
the following:
- publish a Proposed Recommendation
- publish a WG Note
- abandon an unfinished document

i.e. the choice between these end states is made later, and does not 
need to be made on first publication.
If there is clear intent on first publication for one of these outcomes, 
then it should be stated. However, such intent is revocable.

The differences between an informative Recommendation (such as the OWL 
Guide, or OWL Overview) and a Working Group Note are:
- necessary level of consensus
- level of review

i.e. a Recommendation is a better reviewed document, with wider consensus.

Hence, to plan to publish something as a WG Note is to decide up front, 
that we do not wish to dedicate review effort to it, or that we do not 
believe that it is important to get a wide consensus concerning it.

In general, I believe that the bulk of our user facing documents should 
be widely reviewed, and have wide consensus, hence, in general I prefer 
Recommendation track.

However, one thing that happens from time to time, is that a document is 
started, and the enthusiasm of the group or the editors dims, and review 
comments go unaddressed for too long, and the working group partially 
loses confidence that the editor will deliver the quality of document 
(in terms of review and consensus) that is required. In this case, a WG 
Note, is often an explicitly lesser end state.

Particularly given that our charter mandates certain deliverables of an 
informative nature, I think, in general, we should be aiming to make 
them rec track, and only exceptionally as WG Note, when we genuinely 
believe that the higher quality of a rec track document is not worth the 
extra work, or where there is some actual mess-up where some WG 
participants have less time and effort than we initially hoped, and a WG 
Note is an appropriate fall-back strategy.

A further type of document that often is a WG Note, is one that is an 
afterthought, not envisaged in the charter, but the WG becomes aware 
that in some sense it is necessary. In such cases, often the appropriate 
timeline for wide review and consensus is not available, and a WG Note 
is a sensible choice, due to timeliness. At this point in our WG's life, 
I don't believe we should be identifying afterthoughts!

Jeremy
Received on Thursday, 29 November 2007 16:43:53 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:13:27 GMT