Re: help on writing Greek characters
- From: Lee Sau Dan <danlee@xxxxxxxxxxxxxxxxxxxxxxxxxx>
- Date: Sat, 05 Aug 2006 00:28:20 +0800
"Jukka" == Jukka K Korpela <jkorpela@xxxxxxxxx> writes:
Jukka> No single font contains all Unicode characters.
s/character/glyph/
Jukka> "Unicode fonts" are fonts that use Unicode coding (Unicode
Jukka> code numbers for characters), and they should not be
Jukka> expected to cover all of the about 100,000 characters in
Jukka> Unicode. Besides, there are fonts under the same name but
Jukka> with different character repertoires - different versions
Jukka> of "the same font".
If memory serves, a TrueType font may contain more than one encodings.
Internally, the glyphs are named, and there are tables mapping from
character code to glyph names. One table per encoding. Type1 fonts
have only 1 such table, but it is relatively easy to derive new fonts
with different encodings by sharing the glyphs (plus other
information), substituting just another such table. This is called
"recoding" a Type1 font.
Jukka> Thus, even if we assume that all software is
Jukka> Unicode-capable (which surely isn't true), we cannot just
Jukka> use characters and expect everyone to see them.
You would need a font to contains the complete set of glyphs for all
defined characters. Another approach, which is more common, is to
introduce the notion of "fontsets". A fontset is a set of fonts,
usually covering different (but can be overlapping) regions of the
Unicode coding space. In this way, it is possible to construct a more
complete coverage out of a set of smaller but readily available fonts.
But it is not easy to find a set of fonts which are visually
consistent in design and style.
Jukka> I'm worried about Google Groups as an archive that people
Jukka> use - for good reasons. When our messages have expired from
Jukka> most news servers, typically in a few weeks, they will
Jukka> continue their life indefinitely on Google Groups. This
Jukka> means that non-ASCII characters will often appear as
Jukka> transmogrified into something else.
But Google's interface (harnessing a web browser) is not the only way
to read Google Groups messages. At least, Gnus/Emacs has an 'nnweb'
backend, which allows one to read Google Groups messages inside Gnus.
I haven't used it, though.
Jukka> Thus, there are strong reasons to stick to US-ASCII on
Jukka> Usenet.
It depends on what groups you're reading/posting to. e.g. in de.*
newsgroups, it would be considered preferable to use ISO-8859-1 or
-15, if not UTF-8. For hk.* newsgroups, BIG5 (or even BIG5-HKSCS) is
the norm.
--
Lee Sau Dan 李守敦 ~{@nJX6X~}
E-mail: danlee@xxxxxxxxxxxxxxxxxxxxxxxxxx
Home page: http://www.informatik.uni-freiburg.de/~danlee
.
- Follow-Ups:
- Re: help on writing Greek characters
- From: Jukka K. Korpela
- Re: help on writing Greek characters
- References:
- help on writing Greek characters
- From: elagabalus
- Re: help on writing Greek characters
- From: Andreas Prilop
- Re: help on writing Greek characters
- From: mb
- Re: help on writing Greek characters
- From: Geoff
- Re: help on writing Greek characters
- From: mb
- Re: help on writing Greek characters
- From: Geoff
- Re: help on writing Greek characters
- From: Peter T. Daniels
- Re: help on writing Greek characters
- From: Geoff
- Re: help on writing Greek characters
- From: Peter T. Daniels
- Re: help on writing Greek characters
- From: Jukka K. Korpela
- help on writing Greek characters
- Prev by Date: Re: help on writing Greek characters
- Next by Date: Re: help on writing Greek characters
- Previous by thread: Re: help on writing Greek characters
- Next by thread: Re: help on writing Greek characters
- Index(es):
Relevant Pages
|
Loading