From: "John Cowan" <[EMAIL PROTECTED]> > In particular, is it sufficient that the character strongly resembles an > existing character or combination of characters, but for one or another > reason needs to be distinct from it?
I see no good reason, unless such distinction existed in a legacy encoding with which Unicode wants to maintain a roundtrip (bijective) compatibility -- see many examples in Chinese and Arabic. But now that all popular legacy encodings have been transcoded to Unicode, there will be little use of such compatibility decompositions which are quite a nightmare, unless encoding problems are discovered later when a script will be extended to support more languages, where an initially unsuspected semantic difference will be needed by encoding a new character with a compatibility decomposition.
