W3C home > Mailing lists > Public > w3c-rdfcore-wg@w3.org > April 2002

containers

From: Jeremy Carroll <jjc@hplb.hpl.hp.com>
Date: Wed, 10 Apr 2002 18:11:59 +0200
To: <w3c-rdfcore-wg@w3.org>
Message-ID: <MABBLGKMPIJFCKFGDBEPGEKCCAAA.jjc@hplb.hpl.hp.com>

I attach a document about containers.

As I understand it, we have closed nearly all the container issues

...

I was considering one of the open ones,

( rdfms-seq-representation )
...

but I got a bit carried away.

I seem really to have ended up addressing a lot of closed ones, particularly

   rdf-containers-otherapproaches

Anyway.

In short, I think the end marker is not a good idea, because it is making
the current mess of containers even messier (I don't really provide
arguments for this yet).

The specific suggestion I make is that closed bags should be indicated with
a size property taking a string value understood as an xsd:integer.

I attach what I think is a clean model of containers. Either we should put
it in the model theory, as a clarification of M&S (which I genuinely think
it is - it is *only* model theoretic, not a change to the graph) or we
should at least make whatever changes we are going to make to satisfy the
closed container requirement from WebOnt in light of this proposal.

Noting that the attached document really is about a postponed issue, you may
or may not want to look at it.

If you do want to browse the document, I suggest the following reading path
for everyone bar Pat.

Read the first few paragraphs up to the first picture.
Jump to the test cases from the table of contents (part 5).
See if enough of the entailment tests satisfy your intuitions about what
Bag, Seq, Alt in M&S mean.
Then go back and look at the pictures.
Ignore the maths :)

Pat - I was offline when I wrote it and only had the first published MT WD.
I did make a few changes after checking with the latest WD but there might
still be a few leftover bits, which I suspect you can bear with.

If I hear enthusiastic support I may suggest reopening the relevant
postponed issue.

I am not in Bristol today. I'm not much expecting  Brian to be pleased with
me. Fortunately I am spoofing my from-address and the HP firewall doesn't
like people outside HP spoofing an HP e-mail address, so he will never get
this message.

Jeremy



Received on Wednesday, 10 April 2002 12:05:39 EDT

This archive was generated by hypermail pre-2.1.9 : Wednesday, 3 September 2003 09:47:24 EDT