Re: JSON-LD instead of application/microdata+json

It makes perfect sense that microdata use JSON-LD, and this did come up in discussions on the HTML Data Task Force. JSON-LD is a rec-track doument and is a perfect example of representing this kind of information.

That said, the HTML microdata spec conforms to Hixie's world view that it does not make sense to try to provide a universally understood representation of structured data; that the data can only be understood by the specific audience it is authored for. In this case, being able to provide a datatype for, say, schema:dateCreated is not important, because any application using it would already have built-in knowledge about what the datatypes should be.

It's worth persuing further, perhaps this time will make the difference. But, if you don't have very specific reasoning for him for doing this, I suspect that you won't get anywhere.

Gregg Kellogg
gregg@greggkellogg.net

On Aug 9, 2012, at 7:51 AM, Markus Lanthaler <markus.lanthaler@gmx.net> wrote:

> Hi folks,
> 
> you might be interested to join the discussion I started over at the whatwg
> list [1]. I looking at application/microdata+json and saw that it quite
> similar to JSON-LD but lacks IMO a few fundamental things (such as not being
> able to distinguish between a literal and an IRI). I therefore asked if
> JSON-LD was considered as an alternative to defining
> application/microdata+json. It wasn't.
> 
> [1]
> http://lists.w3.org/Archives/Public/public-whatwg-archive/2012Aug/0080.html
> 
> 
> --
> Markus Lanthaler
> @markuslanthaler
> 
>

Received on Thursday, 9 August 2012 16:14:42 UTC