Re: Deprecate in favour of /ns/rdf# ??

On Fri, 2013-11-29 at 10:08 -0500, David Booth wrote:
> On 11/28/2013 10:20 AM, Damian Steer wrote:
> > The issue is that they're hard to remember for authors, mostly, so we'd be better off saying:
> >
> > * rdf and rdfs prefixes are always bound to (whatever they are :-)
> > * rdf and rdfs prefixes may not be bound to anything else
> > * parsers should issue a warning if a document attempts to override them
> . . . with a different URI, I assume you meant.
> >
> > Which is backward compatible, and requires little effort from library vendors.
> Excellent, practical suggestion!   Amidst all the other discussion, I 
> don't think this suggestion got the attention that it deserved.  I would 
> add one more requirement also:
>   * parsers should issue a warning if rdf: or rdfs: are used with terms 
> that are not defined in the RDF or RDFS specs (respectively).
> That would detect mistakes like writing rdf:label instead of rdfs:label.

Is that the role of the parser or the editor?

While baking this into the parser would help. I think identifying
(writing mistakes) as they are written is best.

On that note: GEdit has been highlighting[1] the correct namespace and
terms for rdf/rdfs/owl for almost five years now.

Anybody interested in helping with the latest round of formats (sparql
1.1, turtle, etc) highlights?



Received on Friday, 29 November 2013 15:49:50 UTC