Re: role of DOM vs. SPEAKROW

I concur with T.V. Raman's general statement that poor design practice
should not be encouraged in the name of accessibility. Nevertheless, a
macro which is specific to an individual table, owing to the fact that it
combines variables and customised text strings, transcends the conceptual
dichotomy between content and style. If a generalised reading strategy is
to be employed, involving a mixture of audio cues, changes in voice
characteristics, pauses, etc., then it should be regulated by a style
sheet, but if the author wishes to offer a custom rendering of a
particular table, which includes both text strings and references to data
cells, then there is a strong argument for either treating at as either
(1) an in-line style; or (2) a kind of description of the table which can
be linked to it via RDF or a similar mechanism. If the former option be
preferred, then all that is needed is a style language of sufficient
complexity, which incorporates the features required for such a rendering.
The latter option would involve the creation of a scripting language for
generating audio descriptions of tables (including speech characteristics,
text strings supplied in the script, etc.). The main problem with the
script approach is that there is no universal scripting language currently
available and, so far as I am aware, none has been or is about to be
endorsed by the W3C.

Received on Wednesday, 19 November 1997 02:56:09 UTC