- From: Robert Burns <rob@robburns.com>
- Date: Wed, 4 Jul 2007 19:53:11 -0500
- To: Sander Tekelenburg <st@isoc.nl>
- Cc: public-html@w3.org
On Jul 4, 2007, at 9:09 AM, Sander Tekelenburg wrote:
> Still, I think both conditional comment approaches are too hard for
> authors.
> Especially if we mean to entice them to use <picture> instead of
> <img>.
> Robert Burns' approach is the only one that seems reasonable to me
> to expect
> from authors. Provided no <param> is needed, the spec's object
> definition
> becomes author-friendly, and IE is fixed. {Ahem} :)
I had another thought about a a long-term solution (and basically
this is about long-term solutions) anyway.. Is it possible that, by
moving to an xml serialization, this problem will be solved? In other
words, can we just do <img src="myimage' >fallback</img> whenever
HTML5 content is served as XML. To me this should work (at least as a
complement to the other solutions). Does anyone have any knowledge of
some test cases already online for this?
IE, would of course be difficult to test, though I believe there are
ways to fool it into processing XHTML as XML.
Take care,
Rob
Received on Thursday, 5 July 2007 00:53:24 UTC