[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Adobe Glyph List



At 02:17 PM 98/02/16 +0100, Thierry Bouche wrote:

>« http://www.adobe.com/supportservice/devrelations/typeforum/glyphlist.txt

>contains some `pearls' like

even better, look at UNICODE itself http://www.unicode.com

>002D;hyphen;HYPHEN-MINUS

This is inherited from the ISO 8859 madness. They decided to make
char code 45 a `minus,' when everyone needs a hyphen in text.
So while the standard says `minus,' most people use that char code
for hyphen.  UNICODE `compromised'.  Some Lucida fonts have
a `hyphus' :-) for this purpose.

ISO 8859 is a mess in other ways:  96 is `grave accent'
instead of quoteleft, which is much more important.

And 36 is apostrophe-quote, i.e. quotesingle (a `neutral'
upright quote) instead of the more useful quoteright.

Then there is the issue of `circumflex accent' versus `asciicircum'
and `tilde accent' versus `asciitilde'

And in general, the arbitrary inclusion of some accents and exclusion
of others.  Ideally you'd want all the accents needed to construct the
composites -- or none at all.

Windows ANSI inherits this ISO Latin 1 mess, but tries to work around 
it by adding needed characters in the 128 - 160 range...T1 fortunately 
avoids some of these ISO Latin 1 problems while copying large parts.

>2245;congruent;APPROXIMATELY EQUAL TO
>2248;approxequal;ALMOST EQUAL TO

Well, fortunately they have little pictures to help disambiguate.
It also shows the problem with two different meanings mapping
to the same glyph (although UNICODE dogma is to have two
entries in this case).  The category of `equal' and `not' in math
yields all sorts of delightful attempts to describe in English:

2247 NEITHER APPROXIMATELY NOR ACTUALLY EQUAL TO :-)

> this leads to the conclusion that unicode names are rarely
>appropriate, and should be used with much care...

How about omega1, which is actually an alternate form of pi :-)?

>also, nothing related with maths, but could someone explain me why OE
>is a ligature, while AE is a letter ?!

>From UNICODE 2.0:

00C6	AE	LATIN CAPITAL LETTER AE = LATIN CAPITAL LIGATURE AE

00E6	ae	LATIN SMALL LETTER AE = LATIN SMALL LIGATURE AE
. IPA
-> 0153 oe latin small ligature oe

0152	OE LATIN CAPTIAL LIGATURE OE

0153	oe LATIN SMALL LIGATURE OE = LATIN ... LETTER O E
. French, IPA ...
-> 00E6 ae latin small letter ae
-> 0276  oe latin letter small capital oe

0276 LATIN LETTER SMALL CAPITAL OE
. low front rounded vowel
-> 0153 oe latin small ligature oe

The last of these being in the `IPA Extensions' section

So they do recognize that both can be viewed as ligatures.  But somehow
AE is `more' of a `letter' and less of a `ligature' than OE :-) . 

regards, Berthold.