Le Ven 6 avril 2007 11:18, Karsten Wade a écrit :
It might help if it was understood what types of the problems that entities traditionally solve for technical writing.
Entities have historically been used mostly to declare non-ascii symbols, which is not necessary anymore in an UTF-8 unicode work
They also work for technical stuff which does not appear in text discourse and is not supposed to be translated (Copyright strings, etc)
Any other use fails in an international context (but happens in mono-locale works)