OBJECT for phonetic (more)

For what Max wants to do, we need to ensure that the kana in the
'data' attribute do not overwrite the Kanji in the content of the
OBJECT element.

The question is whether use of 'class="pronounced"' [or
equivalent] will be enough to get browsers to let the OBJECT
content flow through to the visual display.

This would also serve for what Dave wants to do with ad hoc text
that sounds right when pronounced according to the habits of the
incumbent language (language of the current context).

For the MIME registration of types, we should work in concert
with the linguistics profession.  What would work best for us
would actually be to register things like SAMPA as audio
subtypes, because they are an encoding of sound.  Then it would
be clear that even though they are encoded using text characters,
the OBJECT data are only applicable to the audio channel and the
OBJECT content is what should appear in the visual presentation.

However, for the language-based phonetic scripts, whether kana or
Oxford English Dictionary, it would seem that
'class="pronounced"' and 'lang' are going to have to control the
relative disposition of 'data' and content of the OBJECT.

-- Al

Received on Sunday, 23 November 1997 00:58:29 UTC