[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

Re: Talking of HTML.... Anyone like lock-in?

• To: Bjoern Hoehrmann <derhoermi@g...>
• Subject: Re: Talking of HTML.... Anyone like lock-in?
• From: Bob Foster <bob@o...>
• Date: Thu, 01 Sep 2005 23:11:04 -0700
• Cc: davep@d..., xml-dev@l...
• In-reply-to: <eubfh11k6nhms9ikv3c0p98p24nqmrvutf@h...>
• References: <1123780181.3496.27.camel@marge> <42FB89EB.7060207@o...> <eubfh11k6nhms9ikv3c0p98p24nqmrvutf@h...>
• User-agent: Mozilla Thunderbird 1.0.2 (Windows/20050317)

Ok. Here's the full text of my comment and the response I received:

=============
Subject: schemaLocation attribute has no place in XHTML specification

I strongly object to the provision in section 3.1.1 that: "The start tag 
must also contain an xsi:schemaLocation attribute. The schema location 
for XHTML 2.0 is defined to be http://www.w3.org/MarkUp/SCHEMA/xhtml2.xsd."

This is absurd. XML Schema is incapable of expressing any but the most 
trivial constraints in this specification and thus is useless for any 
practical validation of XHTML.

But it would be just as wrong to require a reference to a RELAX NG (or a 
RELAX NG + Schematron, which would be more useful) schema. The XHTML 
committee should not be deciding for XHTML users how their XHTML 
documents are validated, much less specifying a specific schema for the job.

Bob Foster
Bocaloco Software LLC
http://xmlbuddy.com/
=============
On Mon, 01 Aug 2005 05:28:05 +0200, <bob@o...> wrote:

 > I strongly object to the provision in section 3.1.1 that: "The start tag
 > must also contain an xsi:schemaLocation attribute. The schema location
 > for XHTML 2.0 is defined to be 
http://www.w3.org/MarkUp/SCHEMA/xhtml2.xsd."


This turns out to be a typo. We had agreed that the root element MAY 
contain an xsi:schemalocation.

 > This is absurd. XML Schema is incapable of expressing any but the most
 > trivial constraints in this specification and thus is useless for any
 > practical validation of XHTML.


We disagree that it is absurd. XML Schema can express much that is 
useful,  and is a W3C technology, so seems a likely candidate for inclusion.

 > But it would be just as wrong to require a reference to a RELAX NG (or a
 > RELAX NG + Schematron, which would be more useful) schema.


It would be wrong to *require* it, but not to *allow* a reference to 
other  schemas. In fact it would be rather useful, but we are not aware 
of a  standard mechanism yet to allow references to schemas. Perhaps we 
should  define

     <link rel="schema" type="application/rng+xml" href="xhtml2.rng"/>

 > The XHTML committee should not be deciding for XHTML users how their 
  XHTML
 > documents are validated, much less specifying a specific schema for 
the  job.


No, we should be providing mechanisms to allow them to do it if they wish.

Best wishes,

Steven Pemberton
For the HTML WG

• Follow-Ups:
  • RE: Talking of HTML.... Anyone like lock-in?
    • From: "Michael Kay" <mike@s...>

• References:
  • Re: Talking of HTML.... Anyone like lock-in?
    • From: Bjoern Hoehrmann <derhoermi@g...>

• Prev by Date: Public Review of FWSI Functional Elements and Implementation Methodology v1.0 Specs
• Next by Date: Re: Talking of HTML.... Anyone like lock-in?
• Previous by thread: Re: Talking of HTML.... Anyone like lock-in?
• Next by thread: RE: Talking of HTML.... Anyone like lock-in?
• Index(es):
  • Date
  • Thread