Re: WebID default serialization for WebID 2.x

> If every spec would end up making such recommendations for possible different
> serializations, then publishers trying to support different specs may end up
> being forced to publish all possible serializations, which is a pretty high
> burden to push to publishers.

Good point. It applies in reverse to clients, I guess, which is where I am
coming from, but the burden of having to support all possible serializations
would definitely impact publishers in a much more significant way.

> what will happen when you open such a WebID document in a web
> browser? You get Turtle? That would be absolute nonsense, as WebIDs
> can have HTML representations.

> My personal profile is hosted on a static file server as HTML with RDFa.
> Therefore, I do not have the possibility to provide content negotiation on
> this server,

Hosting WebID documents as static resources necessarily implies the possibility
that a browser might find itself confronted with a text/turtle document, that
an automated client might find itself confronted with a HTML+RDFa document and
so on, given the absence of conneg.

> Why do you interpret "MUST provide Turtle" in a way that precludes conneg?

That is my understanding of the use of MUST vs. MAY in the following paragraph:

> A WebID Profile Document is a Web resource that MUST be available as 
> text/turtle [turtle], but MAY be available in other RDF serialization 
> formats (e.g. [RDFA-CORE]) if requested through content negotiation.

IMHO, the above contemplates the possibility for a publisher to only respond
with text/turtle or possibly with a "406 Not Applicable" if anything other than 
text/turtle is requested.

I'm still not comfortable with the idea that clients will, in practice, have to
support all serialization formats but I appreciate how not doing so leads to
more difficult problems. I think I'm coming around to NetID's approach - no
default serialization format, no requirement for conneg. Compared to WebID as
it is today, I think that would mean dropping the MUSTs on text/turtle. 

> I suspect that you think like I did before August 14, 2021 at 20:34ยน.

It does seem the case, yes.

Received on Wednesday, 26 January 2022 12:25:15 UTC