Re: HTML entry point for the RDF Namespace?

On 16/12/2019 17.45, Ivan Herman wrote:
> At the moment, the file is pure HTML, with the vocabulary embedded in JSON-LD

Is that useful?


> Turning it to RDFa is much more work, 


I don't doubt - pipeline or workflow?

I think it is generally way less work because it is done once and the ns
content does not change often.

However, the current HTML document will require duplicate information
for 1) inline and human-visible content 2) JSON-LD in the script block.

If the amount of work to make it RDFa compatible is an actual concern,
I'm happy to contribute and take that off your plate.


> and I am not convinced it would help the legacy applications I am talking about, which may have been created way before RDFa was defined.


I don't know the legacy applications that you are referring to but is it
when a class of consumers omit Content-Type in the request, or perhaps
they only know RDF/XML or Turtle, they may run into trouble with the
HTML response?

AFAIK, the response is entitled to return Turtle if there is no
Content-Type if it seems appropriate. Which is what's happening
currently re:

curl -iL http://www.w3.org/1999/02/22-rdf-syntax-ns

How are the legacy applications handling the Turtle response which may
have been created way before Turtle was defined?

I'd like to better understand in which cases or class of legacy
applications may have a problem with HTML.

The ns in HTML+RDFa will be a welcomed addition. Useful to newcomers so
they can orient themselves prior to encountering the download prompt
with Turtle.

-Sarven

Received on Monday, 16 December 2019 17:36:13 UTC