2017-04-05 21:32 GMT+02:00 Michael Everson <[email protected]>: > It’s wonderful that Mr Verdy opposes my proposal. I must be doing > something right. > > On 5 Apr 2017, at 20:13, Philippe Verdy <[email protected]> wrote: > > > 2017-04-05 18:28 GMT+02:00 William_J_G Overington < > [email protected]>: > > For example, where WOMAN ZWJ ROCKET produces a glyph for a LADY > ASTRONAUT, thus a change of meaning and I think that it went to UTC as > there was a change of meaning but I am not congruently sure of that.. > > > > SQUARE ZWJ CHESSPIECE or CHESSPIECE ZWJ SQUARE produces a CHESSPIECE ON > A SQUARE, thus a change of meaning. > > > > You're right here. The absence of ZWJ clearly means separate symbols > side by side > > Wrong. ZWJ has no particular directional semantics. >
NO! I did nit give any direction. Direc tion is a separate issue (if you mean there the Bidi algorithm) "Side by side" does not prohibit ligatures but this is just like with letters "side by side" where ordering is defined independantly. So I maintain what I replied. The **absence** of ZWJ clearly means separate symbols side by side (minus typographic/styling effects such as joining and **partial** overlays or kerning: this excludes overlays and complex ligatures that are in Unicode treated with separate encodings; partial overlays include kerning, or simple syllabic composition in 2D layouts for Hangul, Kana/Romaji squares, but excludes complex compositions for Hanzi/Kanji which are encoded specifically).
