W3C home > Mailing lists > Public > www-style@w3.org > January 2011

Re: [css3-speech] ISSUE-153 speak: none; usage incompatible with other values of speak

From: Daniel Weck <daniel.weck@gmail.com>
Date: Sat, 22 Jan 2011 13:13:37 +0000
Message-Id: <33A60F71-BB43-4D78-93F6-7C5555781145@gmail.com>
To: www-style@w3.org, fantasai <fantasai.lists@inkedblade.net>, Charles Belov <Charles.Belov@sfmta.com>, Mark Kenny <beingmrkenny@gmail.com>
Hi Mark, interesting points, thanks.

The current draft of CSS3-Speech Module already extends the semantics  
of "display:none" to signify that elements removed from the  
(graphical) layout are to be ignored by TTS renderers (i.e. implicit  
application of speak:none to the entire affected HTML fragment). It is  
also worth noting that an "aural" stylesheet may be authored in such a  
way that when activated (i.e. when a screen-reader or speech-enabled  
browser is processing the page), it not only adds a layer of speech- 
related instructions, but also alters the visual presentation to meet  
the needs of users who are not totally blind.

<accessibility-hat-removed>

I appreciate that the application scope of the CSS3-Speech Module may  
be broader than assistive technologies, and your point about  
multimedia use-cases is a fair one. If we accept a total disconnect  
between the graphical and the aural dimensions, then a fully invisible  
HTML fragment (i.e. not taking part in the page layout, but still  
available as "raw data" in the DOM) could be picked-up and rendered by  
TTS. Conversely, a visible graphical element may not generate any  
speech output at all.

</accessibility-hat-removed>

The problem is that this surely opens the door to some wacky content,  
potentially breaking the familiar screen-reader experience. Is this a  
tradeoff that we can tolerate, or should we adopt a more rigid  
approach to be consistent with accessibility guidelines ? Given the  
forecasted uptake of JavaScript + HTML5 audio/video in the multimedia  
web, I fail to see a strong compelling generic multimedia use-case for  
CSS3-Speech, which is more likely to be used in the context of visual  
impairments and other eyes-free experiences (e.g. in-car driving  
directions).

Any thoughts on that ?
Regards, Daniel

On 22 Jan 2011, at 12:24, Mark Kenny wrote:

> speak:none is not needed, authors can use display:none instead.  
> Personally I think it makes a lot more sense to reuse existing CSS  
> features, especially when the resulting authoring practice matches  
> accessibility guidelines (e.g. a non-displayed visual element gets  
> ignored by a speech processor, and any visible graphical element  
> gets spoken-out). Is there really a compelling reason to keep  
> speak:none ? I can't think of any. Regards, Daniel
>
> It would be useful where you want to present something to visual  
> browsers but hide it from audio browsers.
>
> For example if you had a page with instructions on how to download  
> something: most visual browsers have similar functionality, but I  
> would imagine audio browsers are different and would need separate  
> instructions, therefore the visual ones would be irrelevant.
>
> Not a particularly brilliant example perhaps, but it's at least a  
> real-world one. Personally I think it could be a more useful feature  
> in the future, especially since HTML and CSS are being used more and  
> more for multi-media applications where flash is used currently. Can  
> you see a use for speak:none; in a game for example?
>
> My other objection to removing it is semantic. "Display" doesn't  
> sound like an auditory property. I suppose here we could get into  
> all kinds of mess with something like "render:none;" for both, but  
> maybe that's too far...
>
> Cheers,
>
> Mark

Daniel Weck
daniel.weck@gmail.com
Received on Saturday, 22 January 2011 13:14:13 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 17:20:36 GMT