Re: RDF Issue Tracking: two issues

Thanks Graham. I agree on the API point; my phrasing below was obscure.

What I meant to say was not that API design will help us clarify our
thoughts on this (though it might). Rather, that API design might
benefit from our having a cleare conceptual distinction between the (pure,
mathematical) RDF abstract model and the (pragmatic, context-tracking etc
etc) behaviour of RDF databases. 

Some discussion in this area bogs down because our minds easily slip
(without clear argument) from...

(i) "you can't have repeated occurance of the same triple in an rdf model
(citing the set theoretic definitions in m+s)" (plausible as far as it
goes) 

to

(ii) "an RDF database / API should throw an exception or fail to change
state if the same triple is added at different times" (false, IMHO).

Anyhow, here's two more items for the issue list. Further suggestions
(+analysis / summaries) very welcome...

Dan


On Wed, 6 Sep 2000, Graham Klyne wrote:

> Folks,
> 
> I thought I had sent my original message (below) to the list, but 
> apparently not...
> 
> So, at Dan's suggestion, my comments, preceded by Dan's response.
> 
> I'll add a small comment:  While designing APIs is a good way to explore 
> issues, I don't think APIs are really the right way to make issues clear -- 
> I think good document review and editing is the way there.
> 
> #g
> --
> 
> 
> >Date: Tue, 5 Sep 2000 15:11:10 -0400 (EDT)
> >From: Dan Brickley <danbri@w3.org>
> >To: Graham Klyne <GK@dial.pipex.com>
> >cc: danbri@w3.org
> >Subject: Re: RDF Issue Tracking: help needed
> >
> >
> >Hi Graham,
> >
> >Thanks for these; you're right, they're both recurring issues. The
> >mime-typed fragment business I consider a Web architecture bug. The
> >multiple statements thing I find more intriguing. IMHO people tend to
> >conflate 'rdf model' with 'rdf-based web data management system', and say
> >that the latter should throw out recurrent triples (and their
> >contexts...). Making this distinction more clearly (eg. in terms of
> >API: an API to what?) might help defuse that issue.
> >
> >Fancy resending this message (include my reply above if you like) to
> >www-rdf-interest and I'll add to the Issue Tracking page.
> >
> >Thanks for your help :-)
> >
> >cheers,
> >
> >dan
> >
> >On Tue, 5 Sep 2000, Graham Klyne wrote:
> >
> > > Dan,
> > >
> > > I think this is an important activity, as I'm fairly sure there are 
> > several
> > > issues that have been raised concerning the RDF spec that are not captured
> > > in one place.  Two that I recall of the top of my head are:
> > >
> > > 1.  RDF uses URI-references to identify rdf resources.  But the meaning of
> > > a fragment identifier is defined only in terms of the MIME type of an
> > > entity associated with the resource identified by the URI part.  How does
> > > the RDF square up to this?  What is the MIME type according to which the
> > > fragment identifier of an RDF resource identifier is interpreted?  Does it
> > > depend on the RDF resource involved?
> > >
> > > 2.  There is a question whether or not there can be two different
> > > statements with the same subject, object and property.  Most people 
> > seem to
> > > say "no".  I have suggested that this should be allowed because it can be
> > > expressed in reified RDF statements and that there should be a 1:1
> > > correspondence between what can be expressed in an RDF model and its
> > > reification.  I'm not wedded to this idea.
> > >
> > > (Even if these issues have been resolved somewhere, it would be worth
> > > collecting the agreed resolutions.  I think they call for revisions to the
> > > RDF M&S, if only editorial.)
> > >
> > > I'll try to post more issues for you as they occur to me.
> > >
> > > #g
> > > --
> > >
> > >
> > > At 09:03 AM 9/5/00 -0400, you wrote:
> > >
> > > >RDF IG,
> > > >
> > > >As you all know, discussion threads on this list tend to revisit old
> > > >themes, and we're dealing with a rather complex web of overlapping
> > > >problems and puzzles.
> > > >
> > > >I've finally put up a skeletal RDF Interest Group 'Issue Tracking' page as
> > > >an effort towards gathering common issues, strategies and resolution
> > > >proposals from the RDF community.
> > > >
> > > >         http://www.w3.org/2000/03/rdf-tracking/
> > > >
> > > >This is a rough, simple start and needs more content, polish and (above
> > > >all) issue summaries. Rather than sit on it for still longer, I'm 
> > pushing this
> > > >to the list in the hope of suggestions, help and content.
> > > >
> > > >So, please have a look at this. Meanwhile, next job: RDF Schema CR 
> > summary.
> > > >
> > > >More on which another time...
> > > >
> > > >Dan
> > > >
> > > >--
> > > >RDF Interest Group chair
> > > >mailto:danbri@w3.org
> > >
> > > ------------
> > > Graham Klyne
> > > (GK@ACM.ORG)
> > >
> 
> ------------
> Graham Klyne
> (GK@ACM.ORG)
>

Received on Wednesday, 6 September 2000 06:04:04 UTC