Philippe Verdy <verdy underscore p at wanadoo dot fr> wrote: > I admit that the precomposed dotted-I was not really needed as > well, as it has a canonical equivalence with the decomposed > letter.
You know the answer to this: one-to-one mapping with existing character sets designed for Turkish. This was a crucial design goal for Unicode long before the concept of canonical equivalence was formally defined. > As both J and dot-above are already encoded, the stability > policy would require a candidate dotted-J to have a canonical > decomposition EXCLUDED from NFC/NFKC recomposition... This and all subsequent talk about capital dotted J is completely frivolous. There is no orthography that uses dotless j in such a way that the dotted/dotless distinction needs to be maintained in uppercase, the way it does for i in Turkish and Azeri. UTC and WG2 resisted encoding dotless j for many years, when it was considered just a glyph form to which acute accents and such could be added, and only agreed to encode it within the past six months when it was shown to have specific uses in mathematical contexts and in a particular Latin transcription of a language normally written in Cyrillic. -Doug Ewell Fullerton, California http://users.adelphia.net/~dewell/
