Peter Constable wrote:
> On 06/04/2001 06:54:40 AM Marco Cimarosti wrote:
> One might argue that, in this situation, it's still legible:
> i 3i italiani / le 3e italiane.
No, these forms would never be recognized as "terzi" or "terze" in Italian.
The superscript property is not optional as with the corresponding English
suffixes "st", "nd", "rd", "th".
Out of the context of this message, I would not have understood what
"Three-I Italians" was supposed to mean. The case of "3e" is even worse
because, in a technical context, it could be mistaken with hexadecimal or
exponential numbers.
Of course, there are several other ways of writing ordinal numbers with only
ASCII characters. For instance, Roman numerals are quite common (especially
for some kinds of things: centuries, chapters, kings, popes). It is also
quite common to simply spell out the numbers in letters: "primo, secondo,
terzo, ...".
But for some other kinds of things, "Arabic" digits with superscript
indicators are more common. One such cases is the numbering of periodical
events such as exhibitions, championships, meetings, etc.
Two of these periodical events are particularly interesting to this
discussion: "olimpiadi" (Olympics) and "campionati" (championship), which
are respectively feminine plural and masculine plural and, thus, would
require the plural indicators with the numbers.
> >4) Unify the masculine indicator with Unicode 3.2's U+2071
> "SUPERSCRIPT
> >LATIN SMALL LETTER I" and only propose the feminine indicator
> (tentatively:
> >*U+2072 "SUPERSCRIPT LATIN SMALL LETTER E").
>
> Yes, I'd do that.
I am in favor economy as well, as far it does not cause ambiguity.
> >(Contra: the U+2071, used as a math symbol, should be
> coherent with U+207F
> >"SUPERSCRIPT LATIN SMALL LETTER N" but, used as an ordinal indicator,
> should
> >be coherent with U+00AA and U+00BA, hence a possible
> problems for fonts).
>
> What do you mean by "coherent with"? I don't see any reason why 2071
> couldn't be used as a math symbol, as an ordinal indicator in
> Italian, or
> as a phonetic symbol. If you are talking about appearance
> (you want the
> ordinal indicators to have a line underneath the letter), that is
> definitely a display issue: not all fonts do that for 00AA
> and 00BB, and a
> font feature could be used in a given font to turn that on or off.
Yes, you got my point. (BTW, my "coherent" was a false friend: I meant
"consistent").
Not in all fonts U+00AA and U+00BA are underlined, but in some fonts they
are. In these fonts, also the glyphs for U+2071 (SUPERSCRIPT LATIN SMALL
LETTER I) should be underlined, for consistency.
But the same character can*not* be underlined when it is used as a
mathematical symbol (I am not a mathematician, but I am quite sure that it
would mean another thing if underlined).
You are correct saying that font features could solve the ambiguity, but
this could turn out to be a chicken-egg problem.
I am proposing two *compatibility* characters in order to maintain a certain
font difference in *plain* text. But, if I am in plain text, I cannot turn
on or off "font features" for portions of text.
On the other hand, if I am in rich text I can do such things, but I don't
need these compatibility characters any more: I can use the real
"superscript" font property.
_ Marco
This archive was generated by hypermail 2.1.2 : Fri Jul 06 2001 - 00:17:18 EDT