On Mon, 29 Oct 2018 12:20:49 -0700 Doug Ewell via Unicode <[email protected]> wrote:
> Richard Wordingham wrote: > > I think this is one of the few cases where Multicode may have > > advantages over Unicode. In a mathematical contest, aⁿ would be > > interpreted as _a_ applied _n_ times. As to "fⁿ", ambiguity may be > > avoided by the superscript being inappropriate for an exponent. What > > is redundant in one context may be significant in another. > > Are you referring to the encoding described in the 1997 paper by > Mudawwar, which "address[es] Unicode's principal drawbacks" by > switching between language-specific character sets? Kind of like ISO > 2022, but less extensible? More precisely to the principle. What is an irrelevant, optional feature in one writing system may be significant in another. I'm currently trying to work out the rules for writing Pali in the Sinhala script - I have to worry about the difference between touching letters and conjuncts. A simple ISCII-like encoding for Sinhala Pali would delegate such matters to the font. Richard.
