Re: Undefined types in XHTML Modularization Schemas

Alexandre Alapetite wrote:

> I suggest a new correction to this problem, in the "module of common
> content models for XHTML 1.1" 
> [http://www.w3.org/TR/xhtml-modularization/SCHEMA/xhtml11-model-1.xsd],
> adding a reference to a group defined in the events module 
> [http://www.w3.org/TR/xhtml-modularization/SCHEMA/xhtml-events-1.xsd].

That sounds promising. I've updated <http://schneegans.de/sv/> with
your modification, and it seems to work. Thanks for your effort; I
tried to modify the schemas myself for ten minutes or so, but got
completely lost.

> Here is an XHTML Basic example 
> [http://www.w3.org/People/mimasa/test/schemas/xhtml-basic-sample.xml]
> that is not working with the current XHTML Basic XML Schema
> [http://www.w3.org/TR/xhtml-modularization/SCHEMA/xhtml-basic10.xsd]
> (2006-02-13).

Yes,
<http://schneegans.de/sv/?url=http://www.w3.org/People/mimasa/test/schemas/xhtml-basic-sample.xml>
gives a strange error message. The problem is obviously located in
<http://www.w3.org/TR/2006/PR-xhtml-modularization-20060213/SCHEMA/xhtml-basic10-model-1.xsd>:

  <xs:group name="xhtml.HeadOpts.mix">
    <xs:choice>
      <xs:element name="meta" type="xhtml.meta.type"/>
      <xs:element name="link" type="xhtml.link.type"/>
      <xs:element name="object" type="xhtml.object.type"/>
    </xs:choice>
  </xs:group>

  <xs:group name="xhtml.head.content">
    <xs:choice>
      <xs:element name="title" minOccurs="1" maxOccurs="1" type="xhtml.title.type" />
      <xs:element name="base" type="xhtml.base.type" minOccurs="0" maxOccurs="1" />
      <xs:group ref="xhtml.HeadOpts.mix" minOccurs="0" maxOccurs="unbounded" />
    </xs:choice>
  </xs:group>

This looks flawed, since it allows

  <head>
    <base href="" />
  </head>

or

  <head>
    <meta http-equiv="Content-Language" content="en" />
  </head>,

but prohibits

  <head>
    <meta http-equiv="Content-Language" content="en" />
    <base href="" />
  </head>.

> Since XHTML Modularisation is still a "Proposed Recommendation",

Given the definition in
<http://www.w3.org/2005/10/Process-20051014/tr#RecsPR>, I doubt that
"still" is the proper word here.

-- 
<http://schneegans.de/>                                              |

Received on Monday, 5 June 2006 17:23:48 UTC