Re: representing deserialized RDF - N3, JSON-LD, RDFa, XML/RDF

On 11.12.13 00:20, Ruben Verborgh wrote:

> What do you exactly mean with “keep”? Should use at all points, or 
> offer conversion mechanisms?
> 
> The former would be difficult, as the choice for representations is 
> something library-specific (as you can see in pretty much any 
> language).

> However, the latter could be cool: if there was some common memory
> format we could agree on; then each library should just convert from
> and to it, and then interoperation is possible (although costly).

ACK, I doubt a bit that we can agree on one common format, even if I
would like to have one. But having a common conversion format would
facilitate things at least.

> Stating the obvious: could JSON-LD play that role?

Probably, for me it's important that we keep prefixes in the game for
this format. Which means when I parse a turtle file with prefixes they
have to be saved in whatever format we agree on and when I serialize to
a format which supports prefixing, they get re-used. I also work with
some common prefixes in my code and I want to be able to use them all
the time.

IMHO Linked Data needs prefixes, otherwise it is really not that much
fun to work with.

regards

Adrian

-- 
Adrian Gschwend
@ netlabs.org

ktk [a t] netlabs.org
-------
Open Source Project
http://www.netlabs.org