• To: John Cowan <jcowan@r...>
• Subject: Re: [Fwd: The problems with Xlink for integration langu ages]
• From: Jeni Tennison <jeni@j...>
• Date: Tue, 17 Sep 2002 12:26:14 +0100
• Cc: tpassin@c... (Thomas B. Passin), xml-dev@l...
• In-reply-to: <200209171115.HAA12780@m...>
• Organization: Jeni Tennison Consulting Ltd
• References: <200209171115.HAA12780@m...>
• Reply-to: Jeni Tennison <jeni@j...>

John,

>> Not if you want to support longdesc as well, I believe?
>
> Thankfully, longdesc seems to be history in XHTML 2.0.

OK, perhaps we can abstract out the general point. There are elements
in certain markup languages that have several attributes that contain
URIs; the different attributes point to things that have different
relationships with the local resource that is the element. In XLink
these count as extended links, and the different locations have to be
represented by separate elements, with arcs defined between them.

In XHTML 2.0, examples are <object> then, which has four attributes
that contain URIs:

  - archive (which actually contains a list separated by commas (!))
  - classid
  - codebase
  - data

And <area>, which has two attributes that contain URIs:

  - usemap
  - href

These kinds of elements make it impossible (I think?) to do a simple
AF mapping between the markup language (in this case XHTML) and XLink.
  
Cheers,

Jeni

---
Jeni Tennison
http://www.jenitennison.com/

• References:
  • Re: [Fwd: The problems with Xlink for integration langu ages]
    • From: John Cowan <jcowan@r...>

• Prev by Date: Re: [Fwd: The problems with Xlink for integration langu ages]
• Next by Date: Re: Don't Let Architecture Astronauts Scare You
• Previous by thread: Re: [Fwd: The problems with Xlink for integration langu ages]
• Next by thread: Re: [Fwd: The problems with Xlink for integration langu ages]
• Index(es):
  • Date
  • Thread