Re: WebOnt use-case discussion format, FYI

On Tue, Apr 27, 2004 at 10:53:30AM +0100, Seaborne, Andy wrote:
> 
> Dan,
> 
> Not having been involved on WebOnt, I can only guess, but the community from
> which WebOnt members came, has, at its core, a set of people who had a
> common language and understanding (not necessary agreement!) from previous
> work.  The community from which DAWG members is more diverse.

Yes, and that makes me nervous for other reasons, as well. But
-shrug-, I guess.

> Now we are moving to a working draft, I would expect that we can make rapid
> progress as there is something concrete and bounded to comment on.  I would
> be inclined to go with the structure that the editor already has, although
> adding the "Goals" section could be useful, but not to delay a published
> draft.

This week I'm reading and thinking about as many Use Case and
Requirements documents as I can find from W3C working groups. I like
OWL's a lot. It may be that I can convert my 'benefits' stuff into
something like its Design Goals (and, *perhaps*, move the less well
supported tech requirements into an Objectives section -- I think Dan
mentioned something like in Leiden -- especially if the chair suggests
some kind of ranking based on our straw polling, but even if not I'll
try to do some of that).

Anyway, having precedents is very helpful to me in editing. So thanks.

> I read the goals section of the WebOnt document as stating good principles
> (and clarification with repesect to RDFS) which can be hard to pin down in a
> single use case, but things that have emerged from practice.  For DAWG, that
> is not so much specific query/protocol items as general design - som eis
> already captured in our requirments (e.g. "Queries expressible in a syntax
> that is easily read and written by people.")

Ah, yes, I'm agreeing with this suggestion, Andy.

> In section 2, it says:
> [[[
> The requirements were chosen based on the aspects of the use cases that the
> working group considered most important, while considering the scope of the
> OWL charter and other design constraints. As such, one should not assume
> that OWL will directly support every aspect of the use cases.
> ]]]
> and the last sentence makes a good point.  The work we do enables, not
> completely solves, the usecases.

Yes, again. It may make sense to have language just like that in our
doc.

Kendall

Received on Tuesday, 27 April 2004 08:00:43 UTC