Re: Facebook's Open Graph Protocol

On Wed, Apr 28, 2010 at 9:31 AM, Melvin Carvalho
<melvincarvalho@gmail.com>wrote:

>
>
> 2010/4/28 Dan Brickley <danbri@danbri.org>
>
>
>>
>> On Wed, Apr 28, 2010 at 3:06 AM, Renato Iannella <renato@nicta.com.au>wrote:
>>
>>>
>>> On 28 Apr 2010, at 02:57, Toby Inkster wrote:
>>>
>>> > If they'd reused existing vocabs, they probably wouldn't have been able
>>> > to keep their data structure as flat as it is. This flat schema may
>>> > prove important for adoption.
>>>
>>> I think you nailed it on the head here Toby.
>>>
>>> There is a "reluctance" to reuse existing vocabs for the desire to have a
>>> "flat" structure (one namespace).
>>
>>
>> That was certainly a major consideration. But there's another sense of
>> 'flat' that Toby perhaps had in mind: they didn't really get into RDF-based
>> modelling, and if you look at it as RDF, you see assertions that technically
>> are about a page. So the page has a 'director', or a 'cuisine'. This is
>> quite simple markup, but is hard to map (even with latest OWL) to other
>> vocabs that talk about 'director' of a movie, 'cuisine' of an
>> EatingOrientedFoodVendageEstablishment or whatever.
>>
>> As the one who helped persuade them to use RDFa, I did try to also get a
>> more RDFish model adopted, and suggested a schema with mappings (to vcard
>> and foaf and portable contacts and dbpedia, ...). This simply wasn't
>> possible on last week's schedule, but I'm following up and will keep you
>> posted...
>>
>
> Awesome work Dan.
>
> I'll be the first to admit that the markup has some shortcomings, but even
> if nothing more than this comes out of the SWXG process, this one step
> alone, can be considered to be a HUGE success, for the W3C, Linked Data and
> all of Web 2.0.  Well done!
>
> Credit must also be given to David Recordon, who got up in the early hours
> to speak with us, six months ago, and the rest of the facebook team.
>

Yes it was of course David I've been talking with. I hope we can continue to
help them navigate there way around the SemWeb standards stack. It's not
always easy reading...

Meta question: do we think the "prefix=" variant on xmlns= is stable enough
to encourage people to use?  I see it introduced in the new drafts although
I didn't yet find the detailed spec text.
http://www.w3.org/TR/2010/WD-rdfa-core-20100422/#s_Syntax_overview

On the opengraphprotocol list, some folk are asking whether it's ok to put
xmlns="..." into their non-XHTML HTML pages, and I'm unclear of the answer.
The spec is clearly not yet a W3C recommendation, but sometimes the Web
won't wait. Are parsers supporting prefix= ?

Dan