Richard Wordingham scripsit:

> On the topic of fonts, I was flabbergasted to discover that the Khmer
> alphabet in Unicode needed a very recent version of Uniscribe to work in
> any reasonable fashion. An older version (1.409.2600.1106 of
> usp10.dll, dated 2002) couldn't handle such simple tasks as preposing
> the preposed vowels! What would the problem be here? Where does
> the information that vowels are preposed reside? I had assumed it lay
> in the font - does it reside in Uniscribe instead?

No, in Windows systems everything that is character-specific resides in
Uniscribe rather than the font; this includes which characters are
reordrant. Microsoft's viewpoint is that since this information is the
same for any Khmer font, there is no point in storing it in the font.
Unfortunately, Uniscribe is a monolithic object; fortunately, you can
(AFAIK) always replace an older copy with a newer one without loss of

Evolutionary psychology is the theory John Cowan
that men are nothing but horn-dogs,
and that women only want them for their money.
--Susan McCarthy (adapted) jcowan@...