Re: html 5 and accessibility issue - need of fallback content

Just to address one point : I will address others later.

Lachlan Hunt wrote:

 > * <embed>fallback</embed>
 >   - pro: consistent with <object> fallback.
 >   - con: not backwards compatible at all, embed is an empty element.

Since WHATWG have jettisoned SGML as the foundation on which
HTML is based, and gone on to define a parsing model of their
own, what is to stop that model from specifying that if
an <embed> tag is encountered, the parser is required to
look ahead (honouring nesting) until the first unmatched
</ ...> tag is encountered.  If that tag is </embed>
(case-insensitive), then <embed> is being used as a container
and parsed as such; if it is anything other than </embed>
(case-insensitive), then if it matches the currently open
nest it should close it, otherwise the error-handling actions
should be invoked.

Indeed, given the willingness of the WHATWG to allow
constructions such as "<img ... />, could we not take
advantage of this to mandate that

 <img .../>
 <embed .../>

etc., are closures, whilst

 <img ...> ... </img>
 <embed ...> ... </embed>

etc., are containers ?  If we /were/ to adopt this approach,
I think that many of my reservations about the current
approach would be considerably reduced if not completely
eliminated.

Philip Taylor

Received on Sunday, 1 July 2007 11:00:11 UTC