>The problems come with things like #&x0F68; = &TIBETAN_LETTER_A; versus
>འ = &TIBETAN_LETTER_-A;. Plus, 13791 names of up to 83 characters each
>(U+FBF9 is ARABIC_LIGATURE_UIGHUR_KIRGHIZ_YEH_WITH_HAMZA_ABOVE_WITH_ALEF_
>MAKSURA_ISOLATED_FORM) is a lot of space.
Focusing the discussion specifically to the Arabic character, is
there general agreement that some map containing this
(name,character) pair needs to exist someplace and that not everyone
needs it? I don't need this particular mapping for the documents
that I use regularly, but if I get a document from someone that has
this character name I would like someplace to go to find out what it
is.
Failing the mapping stated in the DTD, I would expect the document to
contain a reference to the mapping, or at least to know that in some
content a particular document describing this relationship considered
the default (i.e., when processing XHTML documents the ° is a
degree symbol without the mapping being explicitly stated in the
document).
Joel
--
