[Date Prev]
| [Thread Prev]
| [Thread Next]
| [Date Next]
--
[Date Index]
| [Thread Index]
Re: [xml-dev] what's missing in XML? What's coming?
- From: Liam R E Quin <liam@w3.org>
- To: David Carlisle <davidc@nag.co.uk>
- Date: Wed, 04 Jan 2012 21:42:36 -0500
[oops, I don't think this went out, sorry if it's a duplicate]
On Tue, 2012-01-03 at 10:28 +0000, David Carlisle wrote:
> On 02/01/2012 17:48, Liam R E Quin wrote:
>
> >> Preferably, build the default mappings of the entities into XML.
>
> > It's interesting that I haven't seen any comments along these lines
> > from the MathML WG.
>
> yes you have:-) we even turned up in person to an xml f2f meeting (at a
> TPAC prior to what became xml 1.1 in the end) asking for use of entity
> names in fragments without an explicit doctype to be well formed in xml 2.
You are right, sorry. It was in 2002.
> Given the situation now, I wouldn't change xml handling in
> application/xml, but I would change it in application/xhtml+xml which is
> currently completely broken: it's not possible (without severely abusing
> FPI identifiers) to have a file that is served with that mime type that
> is both valid according to the xml tool chain and well formed according
> to the xml parsers used by browsers supporting html5 if entity
> references are used.
>
> http://lists.w3.org/Archives/Public/public-html-xml/2011Dec/0019.html
Yes, I agree that's a problem.
The answer might lie in some other syntax; I don't think John Cowan's
utf8+names proposal is practical, and adding the entities to XML would
encourage creation of documents that would break on existing parsers (as
would allowing errors in documents - sorry Henri).
An alternate syntax for symbolic names would work.
Tim Bray years ago proposed using Unicode long names, but (1) that
assumes every character anyone wants is in Unicode, which is manifestly
false - see e.g. [1] for a counterexample... and (2) Tim proposed only
the English names, which isn't going to fly.
An alternative might be e.g.
<math xml:escape="\">now \infinity; is available</math>
(if the WhatWG had gone for distributed extensibility, e.g. via
"unobtrusive namespaces", there would be a conceivable mechanism to make
this -- e.g. \ -- be defaulted in Web browsers... sigh)
Liam
[1] http://www.mufi.info/
--
Liam Quin - XML Activity Lead, W3C, http://www.w3.org/People/Quin/
Pictures from old books: http://fromoldbooks.org/
[Date Prev]
| [Thread Prev]
| [Thread Next]
| [Date Next]
--
[Date Index]
| [Thread Index]