Re: ISSUE-94 (n-ary constucts and RDF): Problem with roundtripping when going from functional-style syntax into RDF and back

Would it be considered a problem to add to the syntax specification a  
restriction that it is illegal to use property names that cannot be  
turned into XML namespace-qualified names? This would, I think, solve  
the tripping problem, i.e., would allow all OWL 1.1 ontologies in the  
structural syntax to be serialisable in RDF/XML. Although this would  
in principle introduce a backwards compatibility issue, I doubt that  
it would cause any problem in practice (in practice all OWL  
ontologies use RDF/XML serialisation).

If we can solve the tripping problem in this way, then we can come  
back to the round-tripping problem.

Ian



On 23 Jan 2008, at 20:42, Peter F. Patel-Schneider wrote:

>
> From: Alan Ruttenberg <alanruttenberg@gmail.com>
> Subject: Re: ISSUE-94 (n-ary constucts and RDF): Problem with  
> roundtripping when going from functional-style syntax into RDF and  
> back
> Date: Wed, 23 Jan 2008 15:54:31 -0500
>
>> So
>>
>> DatatypeProperty(<http://example.com/9>)
>> Individual(f:Dad type(owl:Thing) value(<http://example.com/9>  
>> "nine"))
>>
>> isn't round-trippable. However I never really thought of that as a
>> round trip issue as much as a "trip" issue. You can't even get it in
>> to RDF/XML never mind get it back in the same form.
>
> Which means that it is not round-trippable, which is a ...wait for  
> it...
> round-tripping problem.  (Otherwise, why not round-trip through  
> COBOL?)
>
>> -Alan
>
> peter
>
>
>> On Jan 23, 2008, at 1:58 PM, Peter F. Patel-Schneider wrote:
>>
>>> Not true.
>>>
>>> RDF/XML cannot encode arbitrary RDF graphs.  See
>>> http://www.w3.org/TR/rdf-syntax-grammar/#section-Serialising
>>> for more information.
>
>

Received on Tuesday, 29 January 2008 11:43:31 UTC