W3C home > Mailing lists > Public > public-rdfa-wg@w3.org > March 2010

Re: a 'loose' vocab proposal

From: Martin McEvoy <martin@weborganics.co.uk>
Date: Mon, 22 Mar 2010 02:03:00 +0000
Message-ID: <4BA6CFD4.1070106@weborganics.co.uk>
To: Manu Sporny <msporny@digitalbazaar.com>
CC: RDFa WG <public-rdfa-wg@w3.org>
Hello Manu, Thanks for your reply...

On 21/03/2010 17:44, Manu Sporny wrote:
> On 03/20/2010 05:50 PM, Martin McEvoy wrote:
>    
>> 1, Set the "default prefix"  by using the attribute "vocab"
>>
>> <div vocab="http://www.w3.org/2006/vcard/ns#">
>>      <div about="#fred" typeof="VCard">
>>          <div property="fn">Fred</div>
>>          <a rel="url" href="http://example.com/">  Home</a>
>>      </div>
>> </div>
>>
>> If a Parser encounters nested vocab's treat the first ( parent ) vocab
>> as the "default prefix" . For all child vocab's generate  a pseudo
>> prefix, example: ns1, ns2, ns3 ... etc , this will help avoid generating
>> "invalid" RDF/XML  by switching  the default namespace mid parse.   (
>> Ivans and my concern )
>>      
> Could you elaborate what you mean by "pseudo prefix" and "switching the
> default namespace mid parse"?
>
>    

Sorry Ignore the above don't know how that paragraph got in there ....?  
that's not a problem as such just some of my parsing notes... grrr ....

>> 2. Extending vocabulary terms  using "vocab:"
>> Shane McCarron  also mentioned in a message to the list that the vocab
>> attribute  should be scoped just like xmlns (I hope I understood that
>> correctly),
>>      
> I don't think that's what Shane meant. I believe he meant that if @vocab
> was specified like this:
>
> <body>
>     <div vocab="foo: xyz;">
>     ...
>     </div>
>
>     <p>
>     </p>
> </body>
>    

The above example is not demonstrated in the example Shane referenced here:

http://rdfa.info/wiki/RDFa_Vocabularies

his exact words were

"I feel @vocab should be scoped just as @xmlns is today"

http://lists.w3.org/Archives/Public/public-rdfa-wg/2010Mar/0158.html

I totally agree, anything that  draws RDFa away from using xmlns as a 
way of defining CURIES is a good thing, xmlns: works well, but its a 
mistake.... sorry I *have* maintained that view or quite some time now.

> that the "foo" mapping wouldn't be in the list of mappings when the P
> element is processed. He didn't mean "scoped as if declared by xmlns:
> and used as a prefix for attribute names".
>
>    
>> <div vocab="http://www.w3.org/2006/vcard/ns#">
>>      <div about="#fred"  typeof="VCard"
>>           vocab:me="http://gmpg.org/xfn/11#me">
>>          <div property="fn">Fred</div>
>>          <a rel="url me" href="http://example.com/">  Home</a>
>>      </div>
>> </div>
>>      
> We should think twice before going down this route for a number of reasons:
>
> 1. It would require us to make another xmlns-like proposal in the HTML
>     WG - which is hostile to XML namespaces for a variety of good and
>     not-so-good reasons. We would be creating a very long and disruptive
>     discussion between WHATWG, HTML WG and RDFa WG if we were to pursue
>     this.
>    

I disagree.. prefix scoping is not at all the same, scoping via @vocab: 
would eliminate any discussion and the purpose of  xmlns,  because RDFa 
would an attribute of its own to define and map prefixes and keywords... 
RDFa would be able to define the parsing rules of @vocab and make it its 
own.

> 2. It puts the onus on us to resolve the namespacing issues in HTML5 -
>     which we're already doing for xmlns, but will have to now do for
>     vocab: as well.
>    

There isn't any name spacing behaviour  going on ... just matching 
keywords....

> 3. I don't see what this accomplishes that xmlns: + @vocab-as-the-
>     default-prefix wouldn't accomplish.
  ....

> It seems like the same proposal
>     as saying that @vocab defines the default prefix, xmlns: creates
>     prefix mappings, and colon-less CURIEs are allowed.
>    

I dont have a problem with the above statement I agree to a certain 
extent , but dont you think colon-less curries are yet another action 
xmlms: shouldn't be used for? better to stay avay from the topic in the 
first place I would think ;)

Best wishes.

-- 
Martin McEvoy
Received on Monday, 22 March 2010 02:03:28 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 27 April 2012 04:55:06 GMT