I wrote: > > As Philippe surmised, it is a compatibility character, originally included > > in the Unicode 1.0 repertoire for cross-mapping to existing legacy > > encodings: > > > > Code Page 932: 0x81 0x64 > > Code Page 949: 0xA1 0xA6 > >
Asmus responded: > which just pushes that question forward in time... Or *backward*, depending on which direction you are facing, I suppose. In any case, meditation on the content of a JIS X 0208:1997 table, or a KS X 1001:1992 table, or a Big Five table, etc., is in order when trying to understand these kinds of issues. The doohickeys sitting in the row/cell slots in those standards include all kinds of gorp that wouldn't otherwise be necessary, if Unicode encoding were built from the ground up. The explanation for the single-character encoding of an ellipsis character need look not much further than the explanation, for example, of "iii" (Row 5, Cell 03 in KS X 1001:1992): it fits in a box. And things that fit in a box were what were blitted to the character-based screens by the screen generators in the bad old days before image models and graphical screen displays and sophisticated text renderers. The *interesting* question, in my opinion, is why folks feel impelled to use U+2026 to render a baseline ellipsis in Latin typography at all, rather than just using U+002E ad libitum... --Ken
