Re: Deprecate http://www.w3.org/1999/02/22-rdf-syntax-ns# in favour of /ns/rdf# ??

Dear all
Of course changing W3C RDF namespaces is a silly idea, for all the reasons
written by sensible people before.
I would add two things, and answer to Richard's arguments.
- The RDFa core initial context is an effort by W3C to define default
prefixes which allow most users to ignore the very URI and remember only
prefixes.
- The historical namespaces contain history embedded in URIs in the form of
dates, and that's make RDF look like it is : a language, with embedded
history. Would you change "anthropophagy" to "maneating" because it's
simpler :)


2013/11/29 Richard Cyganiak <richard@cyganiak.de>

>
> On 28 Nov 2013, at 23:10, "Charles McCathie Nevile" <chaals@yandex-team.ru>
> wrote:
> But I object to the idea of using schema.org - among other things, using
> a namespace rooted in a domain you don't control is a terrible idea.
>


> I control neither schema.org nor w3.org, so by that logic using either is
> a terrible idea.
>

Well, I suppose what Charles meant by *you* is that *W3C* (and not him or
you or me), controls w3.org but not schema.org. W3C has a clear and
explicit engagement for stability of URIs, versioning policy etc, it's a
standard body with a public process. schema.org is controlled by a
consortium of private actors, has no versioning policy, and terms and
conditions are clear that there is no guarantee of stability of the
vocabulary and related services. The whole thing can disappear tomorrow if
Google and al. decide so, and you won't have anything to say, it's written
in the terms and conditions.
Compare ...
http://schema.org/docs/terms.html
"We may modify or terminate the Website, for any reason, and without
notice. We also reserve the right to modify these Terms of Service from
time to time without notice, and you expressly agree to be bound by such
modifications when posted on the Website."

http://www.w3.org/Consortium/Persistence.html
"... the World Wide Web Consortium hosts (MIT, ERCIM and Keio University)
make the following pledge: that as far as they are able, for resources on
the www.w3.org Web site which are declared (see below) to be persistent ..."

> And holding namespaces for RDF fundamentals is a long way outside
> schema.org's mission,
>
> Indeed! RDF and related standards are W3C specifications, full stop.

I have some news for you. May I direct your attention to
> http://schema.org/Property
> http://schema.org/Class
> http://schema.org/domainIncludes
> http://schema.org/rangeIncludes
> http://schema.org/sameAs
> http://schema.org/additionalType
>

Those are OK and indeed necessary because the semantics of schema.org are
not the semantics of RDFS or OWL, as well explained many times by danbri
and others. But thinking that they could replace rdfs:Class or owl:Class or
rdfs:domain etc would be a big mistake. They are something else. You can't
infer from a schema:domainIncludes assertion the same facts you can infer
from rdfs:domain. And it's perfectly OK.


> That's a pretty good start and I'd like to see more.
>

Yes why not a schema:subTypeOf if schema.org considers that the schema;org
type hierarchy does not commit to the rdfs:subClassOf semantics. I'm not
sure that one would be a good idea, but that's off-topic.

> whereas it seems an obvious thing to expect W3C to do.
>
> W3C does a fine job *holding* namespaces for RDF fundamentals. But that's
> not enough. W3C does a terrible job at removing obstacles to adoption that
> were designed into these namespaces in the distant past.
>

Maybe namespaces choice were an obstacle to adoption in the past, but they
were not the main ones, and I don't think they are now. Granted, having
e.g.; rdf:Property vs rdfs:Class is something looking weird to beginners,
but that's an opportunity to explain history, that all those have not
fallen from the sky in a perfect state at some point, but emerged from a
human process. You have similar examples in natural language, and we live
with them, If you have learnt a bit of trigonometry, you have lived with
cos²(x) bein g a shortcut for (cos(x))², although cos is a function, and
for functions in general, f²(x) denotes f(f(x)), not (f(x))².

Best

Bernard


>
> Best,
> Richard
>


-- 

*Bernard Vatant*
Vocabularies & Data Engineering
Tel :  + 33 (0)9 71 48 84 59
Skype : bernard.vatant
http://google.com/+BernardVatant
--------------------------------------------------------
*Mondeca*
3 cité Nollez 75018 Paris, France
www.mondeca.com
Follow us on Twitter : @mondecanews <http://twitter.com/#%21/mondecanews>
----------------------------------------------------------