Re: action: danbri to get guha's summary of rdf feature implementation (fwd)

On Tue, 12 Jun 2001, Dan Brickley wrote:

>
> Here (fwd'd) is Guha's feature usage summary, as solicited/closed per
> action A1 in our last teleconf:
> http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2001Jun/0109.html
>
> Recap: Guha was soliciting implementation feedback from WG members on the
> parts of RDF that they've used. He didn't get much input, so we just have
> a quick summary below. I might add that my experience of RDF deployment
> is similar: grudging use of the built-in container vocab, little use for
> reification as currently specified.
>
> Dan
>
> ---------- Forwarded message ----------
> Date: Thu, 07 Jun 2001 16:41:34 -0700
> From: R.V.Guha <guha@alpiri.com>
>
> > Guha,
> [...]
> > Did they send you much? Can you summarise easily?
> >
> > Dan
>
> Didn't get much. People use the basic triples. Two companies
> use containers. No company uses reification.

I dunno if Guha lost my response, but I've got experimental code that
uses reification - it sprang out of the "context" meeting that Graham K,
Brian and a mob from the ILRT had a while ago. Not using it in
production though.

It's worth adding that I've got NO "grudging use of container vocab"
because at the time I did this stuff (which I've got on my "to-revise"
list come July) I hadn't slapped a parser on top of the API :-/

Date: Fri, 11 May 2001 10:30:16 +0100 (BST)
From: Jan Grant <Jan.Grant@bristol.ac.uk>
To: w3 <w3@guha.com>
Subject: Re: RDFCore WG 2001-05-11 teleconference agenda

On Fri, 11 May 2001 jos.deroo.jd@belgium.agfa.com wrote:

>
>
> [...]
> > A1: ALL       send RDF feature usage information from real-world
> >               applications you know of to Guha,  who will collect
it.
>
> 1. At this moment we don't use RDF M&S 1.0 defined containers
>    as we have enough support via [] descriptions (e.g. DAML lists)
>    and {} contexts (e.g. statement sets).
>    NB [] and {} are N3 (see http://www.w3.org/2000/10/swap/Primer.html)

I use containers. I've got some API support for them. Never had much to
do with N3. I've toyed with support for statement sets and node sets in
an API, but nothing's settled donwn there yet.

> 2. At this moment we don't use the RDF M&S 1.0 reification feature.
>    We understood the need for subexpressions (e.g. to model Horn clauses)
>    and found the rdf:parseType="Quote" giving good support.

I use reification, too. In fact, I've got a "model within a model" class
(I call it ContextModel) which represents a model with RDF as a
collection of reified statements*. It's mostly experimental, but I'm
(very slowly) building a web-of-trust app on top of it. I've also got a
lightweight record-retrieval mechanism built on top of RDF which can use
reification to ask queries about queries - not had much use for it thus
far though.**

> 3. Description-without-about (anonymous node description) is used a
lot.

Same here. It's quite handy to not have to give something a URI.

jan

* Actually, it's a little more complex than that for efficiency reasons,
but that's the gist.

** primarily because I've got better things to do than sit down and
build a UI for it.


-- 
jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/
Tel +44(0)117 9287163 Fax +44 (0)117 9287112 RFC822 jan.grant@bris.ac.uk
Prolog in JavaScript: http://tribble.ilrt.bris.ac.uk/~cmjg/logic/prolog-latest

Received on Wednesday, 13 June 2001 05:09:44 UTC