W3C home > Mailing lists > Public > www-rdf-interest@w3.org > February 2001

RDF:Seq: no way to say "that's all, folks!"

From: Dan Connolly <connolly@w3.org>
Date: Thu, 22 Feb 2001 01:44:44 -0600
Message-ID: <3A94C36C.DE4FA100@w3.org>
To: www-rdf-interest@w3.org
I think I discovered this issue in July last year;
I keep (a) thinking that I've raised it, and
(b) checking the issues list only to find that it's
not there, so I guess I better get this on record...

I was working on

  The case of the next scribe
  http://www.w3.org/2000/07/scribe-stuff/

trying to take a list of meeting records,
each of which records the date and the scribe,
and sort them by date, oldest first, and
choose the next scribe from the first item
of the result.

I was sort of hacking this up with RDF and prolog,
and I discovered that I couldn't inductively
define sort() over RDF sequences because there's
no empty list in RDF. There's no end marker.

I can say
	:x a rdf:Sequence;
		rdf:_1 :a;
		rdf:_2 :b;
		rdf:_3 :c.
and so on, but I can't say "and that's all, folks!
there's nothing else in the list; if something is
in :x, it's either :a, :b, or :c."

Since then, I have pretty much abandoned using
RDF Sequence syntax and the _1, _2 properties,
and I'm just using first/rest/empty.

  http://www.w3.org/2000/07/hs78/lists
  http://www.w3.org/2000/07/hs78/lists.rdf
  http://www.w3.org/2000/07/hs78/lists.html

which evolved into

  rdf:parseType="daml:collection" section of
  DAML+OIL reference description
  http://pride.daml.org/2000/12/reference.html
  Thu, 22 Feb 2001 05:02:13 GMT

which is both a List/first/rest/emtpy schema
and a new parseType for writing them succinctly.

I wonder where this fist on the spectrum from
radical-new-feature-that's-clearly-out-of-scope-for-RDF-core
to little-tweak-that-we'll-add-in-week-3.

-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/
Received on Thursday, 22 February 2001 02:44:46 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:51:48 GMT