Re: parse() and media types

Thomas Steiner wrote:
> Hi all,
> 
>> I was wondering if we might use:
>>  http://www.w3.org/ns/formats/
> Sounds like a good thing to do, if the list can be extended to contain
> 
> http://www.w3.org/ns/formats/microformats
> http://www.w3.org/ns/formats/microdata
> along the already present
> http://www.w3.org/ns/formats/RDFa

agreed, that would be cool - and Ivan thanks for adding.

> With regards to MIME types and file extensions for common RDF
> serializations, those, as far as I can tell from the various specs or
> spec-like documents, are meant to be as follows (where sometimes
> common practice differs):
> 
> RDF/XML: application/rdf+xml (*.rdf)
> Turtle: text/turtle (*.ttl)
> N3: text/n3 (*.n3)
> NTriples: text/plain (*.nt)

Just a quick FYI, RDF/XML is the only one that's registered, but is also 
often served as application/xml

Turtle is often served as:
   text/turtle
   application/turtle
   application/x-turtle

N3 is often served as:
   text/n3
   text/rdf+n3
   text/turtle+n3
   application/n3

and all 3 + nt are often served as:
   text/text
   text/plain

It's quite a pita to be honest! (not mentioning pdf or similar ;))

> This list, as Nathan points out, leaves out embedded RDF/semantics
> like RDFa, micro{data|formats}.
> 
> I think using an extended list based on http://www.w3.org/ns/formats/
> is a good solution. Plus a name change to the API ;-)
> 
> Best,
> Tom
> 

Received on Sunday, 24 October 2010 13:03:00 UTC