W3C home > Mailing lists > Public > semantic-web@w3.org > May 2013

Re: non opaque primary topics

From: Eric Prud'hommeaux <eric@w3.org>
Date: Fri, 17 May 2013 12:43:19 -0400
To: Richard Cyganiak <richard@cyganiak.de>
Cc: Melvin Carvalho <melvincarvalho@gmail.com>, Tim Berners-Lee <timbl@w3.org>, Semantic Web <semantic-web@w3.org>
Message-ID: <20130517164315.GB18667@w3.org>
* Richard Cyganiak <richard@cyganiak.de> [2013-05-17 09:40+0100]
> On 13 May 2013, at 11:56, Melvin Carvalho <melvincarvalho@gmail.com> wrote:
> > However I think that qnames are predominantly (maybe always?) used in the predicate position
> 
> Absolutely not.
> 
>   <#me> a foaf:Person; foaf:basedNear dbpedia:Galway.
> 
> This thread reminds me of a nice idea for a Turtle extension that was proposed by, I think, Henry Story:
> 
>   @prefix ex: <http://example.com/things/>.
>   @suffix ex: <#it>.
> 
> Now you can abbreviate the IRI <http://example.com/things/foobar#it> as ex:foobar.

Neat. Let's look for motivating use cases. The example above:

  @prefix ex: <http://example.com/things/>.
  @suffix ex: <#it>.
  ex:foobar a <Baz> .

could be more tersely written:

  @prefix ex: <http://example.com/things/foobar#>.
  ex:it a <Baz> .

If there were multiple terms in the single namespace with the same
suffix, we'd see some payoff:

  @prefix ex: <http://example.com/things/>.
  @suffix ex: <#it>.
  ex:hub a <BikePart> .
  ex:tire a <BikePart> .
  ex:spoke a <BikePart> .

over:

  @prefix hub: <http://example.com/things/hub>.
  @suffix tire: <http://example.com/things/tire>.
  @suffix spoke: <http://example.com/things/spoke>.
  hub:it a <BikePart> .
  tire:it a <BikePart> .
  spoke:it a <BikePart> .

The other benefit being that visually, ex:hub, ex:tire, ex:spoke are
all grouped by authority and the names are easy to read.

I have a hard time convincing myself that reading <#it> as an IRI is
consistent (for instance, the base can't be resolved when you parse
the suffix directive), but in pathological cases, that can be said of
prefixes as well.


> Best,
> Richard

-- 
-ericP
Received on Friday, 17 May 2013 16:43:51 UTC

This archive was generated by hypermail 2.4.0 : Tuesday, 5 July 2022 08:45:33 UTC