W3C home > Mailing lists > Public > www-rdf-interest@w3.org > February 2004

RE: Human Friendly Trix

From: Danny Ayers <danny666@virgilio.it>
Date: Wed, 18 Feb 2004 01:50:59 +0100
To: "Phil Dawes" <pdawes@users.sf.net>, "Jeremy Carroll" <jjc@hplb.hpl.hp.com>, "Patrick Stickler" <patrick.stickler@nokia.com>, <www-rdf-interest@w3.org>
Message-ID: <BKELLDAGKABIOCHDFDBPCEJHFJAA.danny666@virgilio.it>

Apologies for arriving late, the list server seemed to have dumped me. I've
not yet caught up with previous posts, but just read the TriX paper and
found a lot to like. Very nice bone-picking of the other syntaxes. Most of
my general comments would be the obvious ones - we need something but the
cat's already out of the bag with RDF/XML etc.

More specific : I very much like the approach to graph naming. I'd have
probably named the triple elements by their purpose rather than rely on
position, but that's not really here or there. There does seem to be some
(pretty understandable) ambivalence as to whether the syntax should be
human-friendly or not. The approach to qnames seems a bit kinky - I'm not
sure about using 15 characters in the elements to say that a URI is
abbreviated, though I must admit I'm not sure about qnames in content at all
(XSLT has a Get Out Of Hell Free card). The XPath/XSLT ability is sweet, I
imagine that what standardisation of railway track gauges looks like. I'd
like to see more on the canonicalisation/signing front, that's a real boon.
I'd also like to see more examples of what can be done using the syntax
extension mechanism, so I can make up my mind whether or not it's as useful
as it sounds. For example...

> http://www.phildawes.net/2004/01/trix

...forgive me for being blunt, but the human-friendly TriX document looks
remarkably like striped RDF/XML. That is with the exception of qnames in
attributes, syntax which tends to make the xml-dev folks become blessed with
kittens and the good gentlemen of TAG remember they had dinner appointments.

> I think TriX is a shaping up to be a serious contender for a second
> RDF XML standard.

It definitely has got a lot going for it. I don't know how it will fare in
the space currently occupied by RDF/XML given existing practice, but
irrespective of that, there may well be a big enough gap in between RDF/XML
and XPathable XML for TriX to take up residence. As a side note, I also very
much welcome Dave Beckett's Turtle syntax as a human form of the language,
without the relatively specific extensions of n3.

Received on Tuesday, 17 February 2004 20:04:17 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:07:49 UTC