Re: RDFON: a new RDF serialization

Bruce D'Arcus wrote:
> I'm not so sure I agree with the syntax-is-the-problem tradition. 
> Surely syntax is one issue

Sure, I agree, and said as much---I'm just attacking one problem here. 
RDF has several problems.

> Um ... sure, but what's wrong with N3 or turtle that requires yet 
> another slightly different syntax?

To be honest, I haven't used N3 or Turtle, but now that you point it 
out, they seem very similar to my RDFON. But RDFON does look more like 
the syntax that I (and many other people who use Java, C#, C++, and now 
Adobe ActionScript) am already used to. Plus I like they way it handles 
literals, pointing to a time when a future version of RDF may give a 
better treatment to typed literals than the way they were added on to an 
already strange treatment of literals by RDF. (RDF has never quite known 
how it wanted to treat literals.)

>
> Your example here is no more simple or elegant, and arguably suffers 
> from conforming to JSON:

I don't know about "suffers", but yes, RDFON looks like JSON. I can't 
say that it's better that N3/Turtle from a technical standpoint, but I 
like RDFON better, if only because it looks like what I'm used to.


>
> Isn't there a JS-based RDF parser that can handle N3?

We can write a parser in any language to handle any syntax---I'm just 
trying to find something that's concise, easy to understand, complete, 
and immediately feels familiar to many people.

> Again, I think you misdiagnose the issue. The "Ajax community" don't 
> ignore RDF because of the syntax. They ignore it because a) they think 
> it's not needed to solve their problems (overkill)

Well, the huge hack that RDF has become *is* overkill, and half of that 
overkill is the RDF/XML syntax. Really the only significant way RDF 
differs from JSON is 1 ) URIs instead of strings for property names and 
2 ) identifiers for object instances. But you wouldn't know it from the 
current RDF spec.

Garret

Received on Thursday, 26 July 2007 22:27:58 UTC