Re: dataset and Dataset

Dear Richard,

[I think we both want to be pragmatic and that actually we agree :-)].

> The question is whether the hasXxx/isXxxOf convention is so
> universally accepted that DCAT would do wrong by not following it.

DCAT will not do wrong by not following it, and this is not a principle 
that MUST be followed, but that MAY be followed when there might be an 
ambiguity for some properties.

> This is clearly not the case. DC, FOAF and SKOS, and the RDF and RDFS
> vocabularies themselves don't use a hasXxx/isXxxOf convention.

Hum hum hum! For the record, FOAF is using a mix: e.g. foaf:primaryTopic 
+ foaf:isPrimaryTopicOf (the later being clearer from my point of view). 
Furthermore, ?s foaf:img foaf:Image has also the advantage of not having 
the property and the class names differing only from capitalisation.

Retarding SKOS, as it has been stated earlier in this thread and I hope 
you will not disagree with this, every single person I know had already 
to look at the spec to remember what is the damned direction of the 
skos:broader / skos:narrower properties! So perhaps the hasXXX/isXXXOf 
pattern would have been more suitable here!

> That is fine, but please accept that there are lots of other people
> who *don't* model their vocabularies this way.

Sure.

> It is true that in *some* cases it's very easy to get confused about
> the direction of RDF properties.

Exactly. My original comment was indeed general and I mainly argue that 
this pattern is useful in some circumstances. It is also useful for 
distinguishing the property/class name differing only by a 
capitalization but one can also adopt other methods (e.g. abbreviate the 
property name, see foaf:img).

> I agree with Dave that in DCAT there's not much of a risk of this
> confusion.

The two original properties that trigger this thread are: dcat:dataset 
and dcat:distribution. Your argument is that because a noun is used (and 
because the domain and range are different), people will not be in this 
case confused by the property direction. I can live with this although I 
think dcat:hasDataset and dcat:hasDistribution will for sure remove 
ambiguity. Furthermore, this will enable to not have property and class 
names differing by only a capital letter which seem to be a problem for 
some of the people using DCAT according to what I read earlier.
Best regards.

   Raphaël

-- 
Raphaël Troncy
EURECOM, Campus SophiaTech
Multimedia Communications Department
450 route des Chappes, 06410 Biot, France.
e-mail: raphael.troncy@eurecom.fr & raphael.troncy@gmail.com
Tel: +33 (0)4 - 9300 8242
Fax: +33 (0)4 - 9000 8200
Web: http://www.eurecom.fr/~troncy/

Received on Sunday, 7 April 2013 09:18:27 UTC