Re: More on <CAPTION> element etc

At 14:04 +0100 UTC, on 2007-06-22, Joshue O Connor wrote:

> Sander Tekelenburg wrote:
>> We know there are many, that's why I asked which one in particular you were
>> referring to [...]
>
> Sorry Sander. It was JAWS.

OK, thanks. Any particular version?

[...]

>> Well, if you rely on a GUI browser that ignores label, accesskey, scope,
>> headers, summary, speech/braille CSS, etc. then even though such is provided
>> by the author of the web page, it might never reach the screen reader at
>>all.
>> Or does it?
>
> This is where the issue of support for attributes and elements by
> browser vendors is so important. If the browser doesn't support them the
> screen reader won't be able to use them either.

Right. I was afraid that was the situation.

[...]

>> Are you saying that all screen readers manage to receive every aspect of a
>> web page,
>
> Pretty much. Those that are supported anyway.

I can't follow. Which "those"? Browers? And "supported" by whom/what?

>>> even when they rely on a GUI browser that ignores certain parts
>>> (label, accesskey, scope, headers, summary, speech/braille CSS, etc.) of
>>>that
>>> web page?
>
> The browser doesn't ignore these elements you, if you are a sighted
> user, just may not see them, as such.

I understand that certain things aren't presented visually. But what I'm
talking about is the situation where a GUI browser ignores certain aspects of
a Web page's HTML (possibly because they are irrelevant to a visual
presentation) and therefore the screen reader relying on that browser doesn't
have those aspects of the HTML available either.

I'm still looking for confirmaton that, due to the architecture of screen
readers, this is in fact the situation. Because if it is, that means that at
them moment the burden of making the Web accessible is much more on the
shoulders of GUI browser vendors than most people seem to realise. And it
probably affects the spec too. If the spec says a visual browser is free to
ignore a certain element or attribute that doesn't apply to visual browsing,
then that would contradict that same spec's reason for including the
element/attribute in the first place.

So hopefully this is all a misunderstanding :)


-- 
Sander Tekelenburg
The Web Repair Initiative: <http://webrepair.org/>

Received on Friday, 22 June 2007 16:26:20 UTC