Andreas Fink <andreas 'at' fink.org> writes: > Comparing to other languages, it doesn't make real sense to use > some 8bit encoding (like UTF8) for Arabic.
I respectfully disagree. Arabic is alphabet-based, so 8bit is more than enough, thus much more efficient. As an example, Arabic Linux distros have long used ISO-8859-6 as system charset; actually, unicode is better suited now, as ISO-8859-6 seems to be a bit broken (I suspect the reason being that computing was mostly designed/standardized by western entities) but theoretically, using 2 bytes per character for an alphabet-based language is wasteful. However, we're going farther and farther from the kannel topic there :) -- Guillaume Cottenceau Create your personal SMS or WAP Service - visit http://mobilefriends.ch/
