"Philippe Verdy" <[EMAIL PROTECTED]> wrote: > Ohhh... I admit this is hypothetic for a possible use, but the candrabindu > case is a precedent coming from romanization of non-Latin scripts: what if > there's a combining x above used to interact over a diacritic and mark its > suppression in corrected texts or in documents related to > orthographic/grammatical rules, or simply because it is needed for correct > romanization of some ancient script...
If special rendering rules are needed for romanisation of particular languages there is a facility in OpenType and other "smart-font" formats to include different rules for different languages written with the same script. One could use this to provide e.g different rendering behaviour for Turkish than for other languages written in Latin and I suspect it could be used in many cases of transliteration non-Latin scripts (presuming a particular language was written in that script) Orthographic rules can certainly be handled by features and lookups in smart fonts. Maybe this is the level on which many of these issues should be handled. We only need new characters where it is necessary to make a distinction, or resolve something that would otherwise be ambiguous, in plain text. - Chris
