Thomas Lotze <thomas dot lotze at uni dash jena dot de> wrote: > So can it be summarized that figures (both arabic and latin) actually > come in only one flavor (upper or lowercase), the other being a > variant glyph, and both kinds of roman numerals being encoded for > reasons other than their semantic meaning?
Both kinds of Roman numerals were encoded for one reason only: compatibility with existing standards, as Ken mentioned. > - As for arabic numerals, is there any convention which form (upper or > lowercase) is meant by U+048..U+057, and which requires special > treatment (variant selector)? You cannot apply a variation selector (U+FE0x, or soon U+E01xx) to the ASCII digits to request a different glyph, at least not until Unicode explicitly defines such a variant sequence. > - Should the two cases of roman numerals be distinguished by directly > using the UVs of one form or the other, or is it, for the sake of > consistency, preferrable to use UVs of only one of them (which?) and > the same UVs with a variant selector for the other, the font mapping > those variants to the other UV range? Roman numerals should be encoded using the letters in the basic Latin alphabet (upper- or lower-case). The only reason to use the characters in the range U+2160 - U+217F is to maintain compatibility with East Asian legacy standards. (The forms in the range U+2180 - U+2183 can be useful for paleographic purposes or giggle value.) Again, you are not permitted to invent your own variation-selector sequences. -Doug Ewell Fullerton, California
