[html4all] Object element support

[Leif Halvard Silli]

Robert J Burns 2008-08-16 11.08:

> [...] I think the object element support is further along than
> most appreciate. Last year I had put together some informal
> test pages to see where the current state of support is: [...]


> [1]: <http://lists.w3.org/Archives/Public/www-archive/2007Aug/att-0027/TestObject.html 
> [2]: <http://lists.w3.org/Archives/Public/www-archive/2007Aug/att-0035/TestObjecttWMeasurments.html 


All debates about OBJECT tend to start/end in a debate about how 
to technially get it to work cross browser. But, actually, I think 
there are some conceptional problems to be solved first.

The most obvious problem with OBJECT is not technical. It is 
semantical. For instance: Does OBJECT have a method for offering 
both a long and a short text equivalent?

Once  the semantical problems are solved, perhaps it would be 
simpler to get cross browser /technical/ support? As the the 
problem now partly is to understand what we need OBJECT to do. 
E.g. Microsoft has a problem understanding that thing: [1. The 
IE-blog.]

	"<OBJECT> is not parsed in a cross-browser compatible way 
(parsing stops at the OBJECT, whereas other browsers continue 
parsing all the fallback content and make it available. No support 
for this parsing behavior is planned for IE8; I'll take this 
opportunity to ask for real-world scenarios that can help me 
prioritize this feature."

Some of your tests demonstrate how to use OBJECT with images, and 
many have recommended OBJECT to be used even for plain images. For 
a semantical comparison of those tests, let's look at an IMG with 
@alt and @longdesc.

It strikes me that  @alt and @longdesc represents *augementative* 
authoring. The reader, if he wants, can opt for the more full 
description in @longdesc. He is served a minimum, but can opt for 
more if he wants.  [2. See Jukka Korpela]

Whereas the OBJECT element, in its current shape and form, 
represents the concept of 'graceful degradation'. Or if you wish; 
The YGWYG = You Get What You Get principle.

As Joshue has said, it is *important* to be able to hide things 
for screen readers. But this is close to impossile with the OBJECT.

Examples: To use

	<object data=image >Alt text</object>

does not, unlike the @alt, close to guarantee that the text 
equivalent is short. And if you do

	<object data=image >Short<object>Long</object></object>

then how is the user able to (a) know that the long text exists, 
or to (b) select the long text if he knows it exists? [I don't 
know, are screen reader software in general able to, technically 
select the different nested OBJECTs?

	Ideas and proposals:

I would like to know what the community think about this:

	<object><img src=src alt=alt longdesc=#nextObject >
		<object id=nextObject>[markup]Long
	[/markup]</object></object>

The image would be displayed by default in all browsers (it works 
fine, even in IE6). However, even the long text will by default 
(in current User Agents) be displayed. I think such a method could 
replace the proposed <ALT> element. [3. HTML 5 Wiki.] Because, the 
proposed ALT element is also, by default, visible, though most 
authors would of course hide it. (Hiding the second OBJECT is 
difficult in IE6, but there are workarounds, such as wrapping.)

I also think @longdesc could be allowed as a boolean. And that it 
should be allowed in the OBJECT element. The meaning, as a 
boolean, should be that the next OBJECT contains a long text 
equivalent. (How @longdesc as boolean should work, needs to be 
fleshed out more.) I think @longdesc is the most important thing 
to ask for if we really want OBJECT to be used. For example then 
one could do this:

	<object><video src=video >
	<object longdesc=nextObject >Short text equivalent
	<object id=nextObject >[markup]Long.
	[/markup]</object></object></object>

In the prolongment of this proposal, I could imagine that all 
elements for replaced content (video, audio, img) could contain 
short text equivalents directly, like this:

	<replaced src=src>Short text equivalent.
	</replaced>

For example <video src=src>Short</video>. And that the presence of 
a boolean @longdesc would indicate the presence of an OBJECT with 
fallback content. For example like this:

	<video src=src longdesc >Short text equivalent.
	<object>[markup]Long.
	[/markup]</objecet></video>

Longdesc should continue to take URLs, for long equivalents which 
requires HTTP.

[1] See "Known issues we are not planning to change in IE8" in the 
official Internet Explorer blog.
<http://blogs.msdn.com/ie/archive/2008/04/10/html-and-dom-standards-compliance-in-ie8-beta-1.aspx>
[2] Jukka Korpela on augmentative authoring: 
<http://www.cs.tut.fi/~jkorpela/html/augm.html>
[3] A Better alt, in the ESW Wiki:
<http://esw.w3.org/topic/HTML/ABetterAlt>
-- 
leif halvard silli