From: Philippe Verdy (verdy_p@wanadoo.fr)
Date: Sun Dec 28 2003 - 21:09:32 EST
"Michael Everson" <everson@evertype.com>
To: <unicode@unicode.org>
Sent: Monday, December 29, 2003 1:45 AM
Subject: Re: German 0364 COMBINING LATIN SMALL LETTER E
> Both s and long s are available for use if anyone wants to use them.
> What's the problem?
There's no problem here in Unicode. We were originaly only discussing about
the default decomposition of the german ess-tsett into s+s instead of
long-s+s, which is more precise and should not alter the NFKD decomposition,
but would really better match the actual usage, origin and semantics as a
true ligature. I know it's not a canonical decomposition and it has
orthographic rules associated with it, but the same rules that indicate or
forbids sharp-s could be written in terms of (long-s, s) pairs face to (s,
s).
Now I wonder if or how a renderer could better render the sharp s, if it's
absent, notably because today's modern fonts really make this decomposition
into long-s and s extremely evident in the rendered ligature (which is very
different from the incorrect beta-shaped glyph).
And if one could compose and render a encoded pair <long-s><s> using the
<sharp s> glyph if present, using the same facilities offered to compose
"et" or "ct" or "ſt" (long s + t) or many other candidate ligatures starting
by <long-s> like "ſa" (in "manſarde) "ſb" (in "eſbrouffe" or "iſba"), "ſc"
(in "eſcale), "ſf" (rare), "ſh" (in English "ſhark" or "ſhall"), "ſi" (in
"ſinus"), "ſk" (in "aſk"), "ſl" (in "iſles" in both Old English and Old
French where the unvoiced "ſ" became later a circumflex on the preceding
vowel "î"), "ſu" (occurs sometimes like in "maſure")...
This archive was generated by hypermail 2.1.5 : Sun Dec 28 2003 - 21:50:27 EST