Re: LC: Opposing OWL/XML format

On 29 Jan 2009, at 09:21, Ivan Herman wrote:
[snip]
> From a GRDDL perspective, _if_ the namespace document (say, an XML  
> file,
> including a possibility for XHTML) includes the necessary GRDDL
> mechanism to yield an XSLT transform, then, well, it has it and  
> will use it.
>
> 'GRDDL mechanism' means that the XML file contains the GRDDL  
> attributes to
>
> - generate an RDF on its own right (just as for any GRDD-able XML)
> - the generated RDF contains triplets to get to the XSLT transform for
> the original data that started the process via a namespace URI

/me wonders whatever happened to conneg :) Or RDDDL

> So yes, indeed, I see two different issues here that are  
> interrelated...
>
> 1. Putting the OWL/XML and GRDDL issue aside, the question is _what_
> should be at the http://www.w3.org/2002/07/owl URI. What format (XHTML
> or RDF) and, if RDF, what should that contain.

I prefer text/html.

> 2. Because the _same_ URI is used both for the OWL/XML namespace in  
> the
> XML sense and for the URI prefix for the OWL case, we may hit some
> problems with the clashes indeed
[snip]

> We could also, ehem, ehem:-) reverse our earlier decision on the  
> OWL/XML
> namespace and separate it into another URI. That would mean a clean
> separation for GRDDL processing.

As you can well imagine, that is a huge non-starter for me :) Indeed,  
if this is a consequence of adding GRDDL support, then it strengthens  
by quite a bit my opposition to GRDDL!

Cheers,
Bijan.