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

Re: best way to write triples?

From: Dan Connolly <connolly@w3.org>
Date: 17 Apr 2002 18:23:24 -0500
To: Pat Hayes <phayes@ai.uwf.edu>
Cc: w3c-rdfcore-wg@w3.org
Message-Id: <1019085805.5164.144.camel@dirk>
On Wed, 2002-04-17 at 18:04, Pat Hayes wrote:
> Help.
> 
> I notice that Patrick is using a 'mathematical' convention to 
> indicate triples in his datatyping draft, like this:
> 
> <ex:Jenny, ex:age, "10" >

I'm wary of that sort of thing...
that sort of thing being: a notation that sorta
looks good to the eye, but isn't fomally specified;
in particular, we haven't implemented and tested
a conversion to/from RDF/xml.

We depend on eyeballs to get it right if we go
this way. I don't like to do that. I like
to have the machine help.

> Last time I looked, Ntriples syntax would have that as
> 
> <ex:Jenny> <ex:age> "10" .

er... technicall, that's legal N-triples; but
<ex:Jenny> is a full URI reference; it's not an
abbreviation for something longer. i.e.
if you mean

	<http://example/vocab#Jenny>

then you have to write that in full.


> Jos writes them using an N3 variant:
> 
> ex:jenny ex:age "10" .
> 
> Maybe we should agree on a common publication format? If so, which is 
> it? Guidance, anyone??

If you can use real n-triples without it being
unmanageably verbose, I suggest that. Otherwise,
I'm kinda partial to the "N3 variant."

Note that n-triples doesn't have metasyntactic varables.
i.e. no "if you see AAA BBB CCC. then include XXX BBB CCC."
stuff. We don't have a formal notation for that sort
of rule; don't confuse the reader into thinking that
we do.


> I ask now because I would like to get it right in what might be the 
> final version of the MT document.


-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/
Received on Wednesday, 17 April 2002 19:23:23 EDT

This archive was generated by hypermail pre-2.1.9 : Wednesday, 3 September 2003 09:47:31 EDT