- From: Albert Lunde <Albert-Lunde@nwu.edu>
- Date: Thu, 21 Oct 1999 23:48:25 CDT
- To: www-html@w3.org
> On the whole, it appears like everybody is right on this matter - which
> makes me think that we are all making different base assumptions about how
> the whole thing would work. For example: is a dictionary, or a set of
> dictionaries, going to be used, or is it an algorythm that determines the
> sound from a set of letters using some rules, or is it a combination of
> methods? Obviously, if it's all done in the dictionary, one tag is more
> than enough - but wouldn't a dictionary that includes all the languages the
> user may want in order to browse the Net be just too big?
Related subjects were discussed long ago on the IETF html-wg list, when
internationalization and language tags were being introduced.
I think it was mainly in connection with spelling and hyphenation:
but the conclusion that seemed to be reached from the examples
posed at that time was that one shouldn't expect a uniform
algolrithm or dictionary to work across langages, so that one
of the purposes of language markup was to allow language-specific
huristics to be used for purposes like spell-checking,
hyphenation, text-to-speech conversion, and CJK kanji font
selection.
--
Albert Lunde Albert-Lunde@nwu.edu
Received on Friday, 22 October 1999 00:48:27 UTC