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

Re: [css3-speech] Updated WD for CSS3 Speech

From: Daniel Weck <daniel.weck@gmail.com>
Date: Wed, 27 Apr 2011 02:40:51 +0100
Message-Id: <ABCEF900-2978-42B4-9EFC-7DF317E1DC46@gmail.com>
To: W3C style mailing list <www-style@w3.org>
Dear all,
the publication of an updated draft [1] of CSS3-Speech is a step in  
the right direction, as a number of improvements have been introduced  
[2] since the 2004 version.

Reviewers/commenters, please note that the editor's draft [3] (on the  
development server) has been updated [4] since the official draft  
publication.

We now need to make progress on the remaining issues [5]. I suggest we  
start with the "phonemes" issue [6]. As there has been prior  
discussions [7] about it, we should be able to move swiftly.

No consensus emerged from the aforementioned discussions, but it is  
clear that the proposed 'phonemes' CSS property doesn't play well with  
the principle of content / presentation separation: a text change  
within the HTML document; with no markup alteration; would require  
modifying associated stylesheets, in order for the pronunciation rules  
to remain in sync with the targeted text tokens. This would be a  
fundamental flaw with respect to the HTML+CSS authoring / maintenance  
workflow. On that basis alone, I predict a fair number of formal  
objections against the current proposal. In fact, some CSS Working  
Group members have already spoken disfavorably during conference calls.

I am under the impression that the proponents of the CSS3-Speech  
"phonemes" property argue in favor of it mainly (and pragmatically)  
because of the lack of support for pronunciation instructions in HTML.  
I'd like to point-out that the upcoming EPUB 3.0 open standard paves  
the way towards a more complete and coherent speech-synthesis  
ecosystem, by leveraging several W3C specifications:

# HTML5 + SSML's "phonemes" functionality (using the 'ssml:ph'  
attribute instead of the original SSML element-based markup)

# PLS Pronunciation Lexicons (by proposing a "rel" extension [8] for  
the "link" element in (x)HTML documents)

# CSS3 Speech Module (all currently-drafted features, apart from the  
'phonemes' property which is judged inadequate by the standardization  
committee)

Given the profusion of EPUB publications and the growing trend to  
distribute via web-browsers, this will set a significant precedent.

Furthermore, the HTML Speech Incubator Group has now formally produced  
a report [9] in which they mention the opportunity to collaborate with  
the CSS Working Group. I suggest that we advocate the same position as  
the EPUB folks, in terms of supporting pronunciation rules within  
content documents via PLS and SSML:phonemes (or such like).

I thereby propose to remove the 'phonemes' property (as well as its  
associated @alphabet at-rule) from the CSS3-Speech Working Draft.

Regards, Daniel

[1]
http://www.w3.org/TR/2011/WD-css3-speech-20110419

[2]
http://www.w3.org/TR/2011/WD-css3-speech-20110419#changes

[3]
http://dev.w3.org/csswg/css3-speech

[4]
http://dev.w3.org/csswg/css3-speech#changes

[5]
http://www.w3.org/Style/CSS/Tracker/products/29

[6]
http://dev.w3.org/csswg/css3-speech/#issue-phonemes

[7]
http://lists.w3.org/Archives/Public/www-style/2010Dec/0237.html
http://lists.w3.org/Archives/Public/www-style/2011Feb/0046.html

[8]
http://microformats.org/wiki/rel-pronunciation

[9]
http://www.w3.org/2005/Incubator/htmlspeech/live/NOTE-htmlspeech.html

On 20 Apr 2011, at 14:19, Bert Bos wrote:

> The CSS WG published an update of the CSS3 Speech module:
>
>    http://www.w3.org/TR/2011/WD-css3-speech-20110419/
>
> The Speech module contains properties for output of documents to
> synthesized speech. They control voice, volume, pitch, speed, pauses,
> etc. They are intended to be compatible with SSML (Speech Synthesis
> Mark-up Language), so that a CSS style sheet can not only be used to
> render a document as speech, but can also serve to convert a document
> into SSML instructions.
>
> The module replaces the deprecated 'aural' media type in the  
> appendix of
> CSS level 2.
>
> The priority of this module has fluctuated in recent years. Although
> some work has been done on it in the working group since the last  
> draft
> of 2004, it never reached a stable enough state to publish a new  
> draft.
> Thanks to a new editor, Daniel Weck of the DAISY Consortium, we may
> finally have enough resources to bring the specification to CR.
>
> Comments are welcome. As usual with CSS drafts, please, send  
> feedback to
> this mailing list, <www-style@w3.org>, and prefix the Subject line  
> with
> the short name of the specification in square brackets:
>
>    [css3-speech]
>
>
>
> For the CSS WG,
> Bert
> -- 
>  Bert Bos                                ( W 3 C ) http://www.w3.org/
>  http://www.w3.org/people/bos                               W3C/ERCIM
>  bert@w3.org                             2004 Rt des Lucioles / BP 93
>  +33 (0)4 92 38 76 92            06902 Sophia Antipolis Cedex, France
>

Daniel Weck
daniel.weck@gmail.com
Received on Wednesday, 27 April 2011 01:41:23 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 17:20:39 GMT