Re: Another way other than @profile, @vocab or @map

On 19/03/2010 15:16, Shane McCarron wrote:
> I have always felt that we need a way to permit the definition of a 
> new default prefix.  In fact, if you look at 
> http://rdfa.info/wiki/RDFa_Vocabularies you can see the wiki page Manu 
> and I started developing about it ages ago.  Toby, you even 
> contributed to it!

Yes I remember that discussion.. It kind of lost momentum as I recall ;)

>
> Actually, looking back, what that proposal did was conflate a couple 
> of things... but I still think it is mostly clean. I updated it to use 
> the attribute name @vocab, because I agree that it is meaningful.  
> Note that in this proposal @vocab can be used to declare a default 
> prefix, declare other prefix mappings, AND implies a follow-your-nose 
> vocabulary extension mechanism.  Also note that in this proposal the 
> target document is an RDFa document that uses the link element with a 
> special @property value to define additional prefix mappings if it 
> wants to.
>
> I am not really trying to muddy the waters here, but I suppose I am by 
> introducing yet another way of thinking about this.  The nice thing 
> about this mechanism is that we could just remove the bits about 
> retrieving a remote document altogether.  It is a way to extend the 
> collection of prefixes and keywords, but it is not really required in 
> my opinion.  Not for the use cases I can think of.

No you are not muddying the waters at all  I belive that Its the same 
discussion almost as the one we are having here.. this example..

http://rdfa.info/wiki/RDFa_Vocabularies#Creating_a_document_that_uses_microformat_terms_via_RDFa

is particularly relevant to this discussion, it should have been solved 
back then in version 1.0 of RDFa as being able to declare a default 
name-space is a basic function of RDF.

anyway +1 for using the url @vocab to declare the default CURIE prefix, 
I would like to add that @vocab should only be used once in a RDFa 
document to avoid invalid RDF in the output document (switching default 
namespaces)

Best Wishes.

Martin.


>
> Martin McEvoy wrote:
>> Hello Toby,
>>
>> On 19/03/2010 13:47, Toby Inkster wrote:
>>> On Fri, 2010-03-19 at 12:21 +0000, Martin McEvoy wrote:
>>>> @profile in this way is behaving just the same as html4 profiles..
>>>>
>>>> "As a globally unique name. User agents may be able to recognize the
>>>> name (without actually retrieving the profile) and perform some
>>>> activity based on known conventions for that profile"
>>>>
>>>> http://www.w3.org/TR/html401/struct/global.html#profiles
>>>>
>>>> In the case of RDFa the "known conventions" would be setting the
>>>> default namespace for the document.
>>> Not quite the same. "name" in the quote above can be translated as
>>> "URI". So when it says:
>>>
>>>     "user agents may be able to recognize the name"
>>>
>>> it means that user agents should only be doing this for URIs that they
>>> recognise. Unless I'm misunderstanding your suggestion, RDFa processors
>>> would be applying the profile as a default prefix whether or not they
>>> recognised the URI.
>>
>> Yes they would be applying the profile as a default CURIE prefix 
>> whether or not they recognise it, ... hmm sounds a little unsafe..  
>> but isn't that what we are suggesting with the rdfa profile proposal, 
>> perhaps I am misunderstanding something ;)
>>
>>> I don't have anything against this general technique - but I don't 
>>> think
>>> it's consistent with the HTML4/XHTML1.x definition of @profile, so a
>>> different attribute would need to be used.
>>
>> I agree (now)  best to avoid @profile and use something new, like 
>> your original proposal, Im glad we discussed its uses first though.
>>
>> http://lists.w3.org/Archives/Public/public-rdf-in-xhtml-tf/2010Jan/0019.html 
>>
>>
>>> A lot of the debate here has been on the syntax and model for profile
>>> documents. Personally I don't think we've had enough debate on whether
>>> profile documents are needed at all
>>
>> I agree ...
>>
>>>   - Martin's suggestion here is not to
>>> define a profile (in the sense that we've been talking about them) at
>>> all, but to just set the default CURIE prefix.
>>
>> Which in my mind is the simplest problem to solve ....
>>
>>> What exactly are the use
>>> cases that show this to be insufficient? Personally, I don't think I've
>>> seen any yet.
>>
>> :)
>>
>> I believe If this group can come to a decision on "how to declare the 
>> default CURIE prefix" a lot of the other problems such as 
>> "prefix-less tokens" and google wanting  to "bundle a bunch of 
>> existing vocabs together" may have agreeable outcome to a certain 
>> extent.
>>
>> @vocab as new attribute name is looking pretty desirable now ;)
>>
>> Best wishes.
>>
>


-- 
Martin McEvoy

Received on Friday, 19 March 2010 15:36:51 UTC