Ivan Herman a écrit :
> Simone Onofri wrote:
>
>> > If I understand the problem properly, there is very little that can be
>>
>>> done here from our point of view. Indeed, afaik, the RSS specification
>>> uses a _restricted_ version of RDF/XML. I think that the fact that my
>>> distiller produces something acceptable by the RSS validators is, though
>>> flattering:-), just a pure coincidence.
>>>
>> Good coicidence :-) and yes, this is the problem. I hope there are few
>> cases with same issue. I'll continue testing on others things to find.
>>
>>
>
> Well, this is an issue of RSS:-( Put it another way, RSS is _not_ a good
> use case for RDFa:-( It is a good example for GRDDL because a
> specialized GRDDL transform for RSS has an exact control over the output
> format.
My 2 cents: it is a good scenario for cascading transforms:
(Transform 1) RDFa -> RDF/XML
(Transform 2) RDF/XML -> RDF/XML RSS restriction
Cheers,
--
Fabien - http://ns.inria.fr/fabien.gandon/