Re: 2 suggestions

On Thu, 28 Mar 2002, Max Froumentin wrote:

> >> http://www.w3.org/2002/03/27-mathml/valid.mml
> >
> > Is application/mathml+xml registered?
>
> No it's not. This is why the w3c server should perhaps not
> send it and use application/xml instead.

OK, then I'll second that suggestion.

>	 Personally, I'd rather
> it being used and recognised by the validator. After all, it is
> suggested by RFC3023, and if there is going to be a MathML mime
> type, it'll be that.

It doesn't need to be explicitly supported.  If the validator just
takes it as XML and accepts a FPI, that's enough.

> > That document has no preamble.  In the absence of a doctype, none
> > of the validators will validate it
>
> It has no preamble on purpose. If the validator gets mime type
> information, or a file extension it should be able to infer that the
> document is MathML.

But your valid.xml does have a preamble, and that is sufficient for
the validators to identify it as MathML 2.0.

Please don't let your WG specify rules about preambles that would be
in conflict with generic XML rules.  We (tool developers) will
heartily curse those who give us conflicting requirements to interpret.

> Yes, I was merely suggesting that upon reporting success in the case
> of the first example, it would be nice to have the validator say that
> it is valid *MathML* and display the valid mathml icon.

We can say the .xml is
  (a) valid XML, and
  (b) MathML 2.0
That's from the FPI.  Adding the MathML icon would be straightforward.

We can say the .mml is well-formed XML, but
  (1) To describe it as mathml, we'd want at least a registered MIME type
  (2) To describe it as valid, we want a FPI.  Without it we have a
      conflict with existing practice and implementations.

-- 
Nick Kew

Site Valet - the mark of Quality on the Web.
<URL:http://valet.webthing.com/>

Received on Thursday, 28 March 2002 14:07:17 UTC