Re: Aural CSS

The problem here is that both your plans defeat one of the main precepts
of CSS: That content and presentation should be seperated. So CSS should
not concern itself with the actual content presented; only the manner in
which it is presented.

Furthermore, if you ask people to write their content twice (e.g. once in
plain text, and once in IPA), they will not do it.

One possibility might be to consider having some way to load the actual
HTML page based on the presentation style present. So you could have one
HTML page in ASCII-IPA, with only aural styles, and another in plain
ASCII, with only visual styles, and possibly without some surpurfluous
content.

In many circumstances, authors may wish to customize their presentation
based on the method being used to convey content to their readers. Or if a
site expects to have a large blind audience they could translate their
pages into IPA for better screen reading. But this behaviour should be
OPTIONAL. By default, all content should be accessible to everyone.

> <<< PLAN A >>>
> 
> Allow the Speak property to take values similar to those taken by the
> Content property (strings, URIs, attr(x)) besides those it already takes
> (none, normal, spell-out).  The element will then be spoken as the
> referred-to text (which would be in a natural language).
> 
> <<< PLAN B >>>
> 
> Allow the Speak property to take a string value which is in a phonetic
> alphabet.  The element would then be spoken according to the rules of that
> alphabet.  Which alphabet is used can be set by means of a Speak-type
> property, which would takes MIME type values.  For example, if the
> Kirschenbaum ASCII-IPA [3] is used, and if it is assigned MIME type text/ipa
> (it now has no MIME type), one could use something like
>    <span style="speak:'\'gEj@t';speak-type:text/ipa">get</span>
> 
> Comments to the list or to me, not both, please.
> 
> Michael Hamm
> BA Math scl, PBK, NYU
> mhamm@gc.cuny.edu
> http://www.crosswinds.net/~msh210/
> 
> ----------
> Notes:
> 
> [1] E.g., that Abbr elements with attr Title set should have the attribute's
> contents read instead of the element's contents (as in the example in the
> text), or that any Strike element should be read as "The quick brown fox
> jumped over the lazy dog").
> 
> [2] As is done at <URL:http://www.crosswinds.net/~msh210/html.html>, e.g.
> 
> [3] Specification at
> <URL:http://www.hpl.hp.com/personal/Evan_Kirshenbaum/IPA/faq.html> with
> meta-info at <URL:http://www.hpl.hp.com/personal/Evan_Kirshenbaum/IPA/>.
>

Received on Wednesday, 21 February 2001 14:40:29 UTC