Re: GRDDL and HTML5

Ian Hickson wrote:
>>> Thus, I posit that 2) is a mis-use of profile="". Certainly, even if 
>>> HTML5 had profile="", it wouldn't license the use of the attribute in 
>>> this way.
>> I would be a mis-use. Is anybody actually suggesting that?
> 
> Suggesting what?

Suggesting to define alternate profiles, just for the purpose of adding 
support for GRDDL's transformation code discovery?

>> Just because "many do not include" doesn't mean it's not used. There may 
>> be code out there that you are not aware of (maybe behind firewalls) 
>> that you potentially break by disallowing the attribute.
> 
> This seems unlikely, since HTML5 isn't yet complete, and so nobody should 
> be using it except experimentally, and HTML4 allows profile="".

Add guess what? That experimental use has shown that the removal of 
profile breaks code, and people speak up.

Sounds like your answer to that is: "come back when HTML5 is done".

>> On the other hand, the price of keeping it is zero (or would have been, 
>> if we would have started with the existing HTML4 vocabulary).
> 
> The price of keeping it is not zero. Just look at the pain it has caused 
> the GRDDL effort. Instead of just automatically supporting all known 

The only pain I'm aware of is caused by the current spec not allowing it.

> vocabularies, the GRDDL team has instead been misled into thinking that 
> having pages declare vocabularies is somehow better.

I can be better, in particular when the community for the vocabulary is 
not large enough for standardizing in a global short-name namespace.

> ...


BR, Julian

Received on Monday, 25 August 2008 12:43:16 UTC