> > And there is (AFAIK) *no* advantage to use UTF-8 > over UTF-16. > > Memory usage for most western languages doesn't > count?
I belive the list had this dicussion before. As I understand it (and, no, I haven't looked at the code myself), the overhead of having linked lists and pointers for AbiWord's internal representation is enough that it doesn't really matter if a single character takes up one or two bytes. If memory usage of the internal table which stores characters becomes an issue, one can always implement something with zlib which can drastically reduce the memory usage. Yes, this will make accessing the data in question more difficult, but, then again, so will using UTF-8. The main purpose of UTF-8 is to allow systems which assume 8-bit ASCII, such as UNIX filesystems, to use Unicode without needing to rewrite the code in question. My vote: Keep on using UCS-4; it is the most simple to deal with, and the memory usage is a non-issue. - Sam _________________________________________________________ Do You Yahoo!? La emoci�n e intensidad del deporte en Yahoo! Deportes. http://deportes.yahoo.com.mx
