At Sun, 28 Sep 2008 10:45:00 -0700, Andreas Raab wrote:
If you need to retain these extra information, sending the strings without going through UTF-8 conversion makes more sense.
Or provide it via additional attributes. I still think that language information would best be modeled by a text attribute - in which case we have a plain Unicode implementation for strings as well as the ability to provide the disambiguation in text where required.
Well, sure, for the more work and more clearner approach. That is what I've been mentioning time to time. The consequence would be that a bare character object or string object won't show up in the proper way; but it is not a big problem.
-- Yoshiki