W3C home > Mailing lists > Public > public-html@w3.org > August 2007

Re: Marking up alternative versions of content

From: Robert Burns <rob@robburns.com>
Date: Sat, 4 Aug 2007 00:22:09 -0500
Message-Id: <2E2930E6-A922-4F5E-855A-1CCF1F0D98D4@robburns.com>
Cc: public-html@w3.org
To: Jason White <jason@jasonjgw.net>


On Aug 3, 2007, at 11:44 PM, Jason White wrote:

>
> I agree with all of Rob's remaining points, including his separate  
> post
> suggesting a flattening of the <object> hierarchy of fallbacks by  
> allowing
> <alt> to be used instead, with appropriate @for references.

One of my suggestions was to continue using:
    [primary --@longdesc(URL)--> fallback]
rather than
    [fallback --@for(IDREF)--> primary]
that your examples exhibit.

> Since the proposed content model of <alt> is transparent, the  
> following
> alternative to @longdesc is possible:
>
> <p><img src="compleximage.jpg" alt="label" id="compleximage">
> <alt for="compleximage"> <a href="description.html" title="Detailed  
> description
> of image">Description</a></alt>
>
> This has the further advantage that the UA behaviour associated  
> with a link,
> unlike @longdesc, is well defined.

It may be too well-defined. In other words it is defined as a  
hyperlinking mechanism to a related resource (rather than a linking  
mechanism to an equivalent / alternate / fallback resource.

However, the same thing could be accomplished with simply

<p>
	<img src="compleximage.jpg" alt="label" longdesc="#compleximage">
	<alt id="compleximage">Description
	</alt>
</p>

OR

<p>
	 <img src="compleximage.jpg" alt="label" longdesc="#compleximage">  
[...]
</p>
<alt id="compleximage">
	<p> [multiple paragraphs of long description]
</alt>

or as I was suggesting with a switch:

<p>
	 <img src="compleximage.jpg" alt="label" longdesc="#compleximage">  
[...]
</p>


[then located immediately after the IMG or P or in an entirely  
different document:

<fallback id="compleximage">
<alt>
	<p> [multiple paragraphs of long description]
</alt>
<alt>
	<object />
</alt>
<alt title='something for a UA to display in the selection mechanism'' >
	<object />
</alt>
</fallback>

The FALLBACK element's content model is thus one or more ALT  
elements. It would need an @id set to be able to reference it with  
the @longdesc attribute. The FALLBACK element could be used within  
OBJECT, VIDEO, AUDIO, and CANVAS to flatten the alternatives, as  
James called it.

> The author also has the flexibility to
> insert the description inline:
> <p> <img src="compleximage.jpg" alt="label" id="compleximage">  
> [...] </p>
> <alt for="compleximage"> <p> [multiple paragraphs of long  
> description] </alt>
>
> An assistive technology, voice browser, etc., could give the user  
> the option
> of reading the description immediately after the image, or after  
> the remainder
> of the containing block-level element (a paragraph in this case)  
> has been
> read, or at the end of the document etc. The user agent/assistive  
> technology
> could also allow the description to be transformed into a link, as  
> in the
> preceding example - in fact, an XSLT style sheet could do this on  
> the client
> or server side, according to the user's preferences, and I'm sure  
> this could
> easily be scripted.
>
> Having listened to numerous audio books with footnotes and figure  
> descriptions
> included, it is clear that the inability to adjust the order in  
> which these
> components are read, or to suppress them entirely during the reading,
> constitutes a substantial limitation that reduces one's  
> comprehension of the
> material. This is another reason why "side-by-side" delivery of  
> alternatives,
> associated only implicitly with the corresponding media objects, is
> insufficient, and why it must be possible for the author to signal,  
> in the
> markup, what is an alternative for which element.

Agreed. Some authors may want to provide complete descriptions in the  
main portion of the document. However for those authors that do not  
want to, we have to provide a mechanism to make the alternates  
available but not in your face. Encouraging/recommending/requiring  
UAs to provide a common mechanism to select among alternates is q win- 
win for both users and authors.

Take care,
Rob
Received on Saturday, 4 August 2007 05:22:20 UTC

This archive was generated by hypermail 2.3.1 : Monday, 29 September 2014 09:38:48 UTC