[Slightly OT] UTF-8 support in *nix terminals (Was: Re: Syntactic differences within parts of speech)
From: | H. S. Teoh <hsteoh@...> |
Date: | Friday, September 1, 2006, 22:01 |
On Fri, Sep 01, 2006 at 05:16:53PM -0400, Henrik Theiling wrote:
[...]
> Hmm, my normal news reader does not support UTF-8 via the xterm terminal
> I am currently using, so I will try the web form to post this:
>
> '買い取る'
> ka i to ru
>
> The first Kanji is the same as in <kau> '買う' and '取る'
> is 'to take'.
[...]
I'm pleasantly surprised that I'm actually reading those characters in
my console-based emailer (mutt). I've recently found a good Unicode
bitmap font for X11, and finally made the switch from xterm to
rxvt-unicode, and switched my default locale to UTF-8. And ever since,
I've been pleasantly surprised, on more than one occasion.
If you're running Linux, I highly recommend installing efont,
rxvt-unicode (or uxterm), and setting your locale to UTF-8 (e.g.
en.UTF-8 or de.UTF-8). Surprisingly many console apps can display UTF-8
correctly if the locale is set right (including most, if not all, of the
system utilities). *Inputting* UTF-8, of course, is a different story,
but at least you can read the stuff.
(I personally prefer rxvt-unicode over uxterm because you can specify
multiple fonts in rxvt, and it chooses the first that contains a
particular symbol. This is useful for selecting language fonts in order
of preference, instead of having to live with ugly glyphs from the one
and only font that actually covers a significant portion of Unicode.)
A funny sidenote: after my recent switch to UTF-8, I started actually
seeing Cyrillic characters in mail, and had a strange paradoxical
feeling of simultaneous indignance and joy one day when I received a
Russian spam mail and actually could read and understand what it said.
:-) (It used to only show up as a bunch of ???'s because I was on the
us-ascii locale which didn't have cyrillic characters.)
<ObConlang>
Speaking of fonts, has anyone experimented with purely vertical writing
systems? (I.e., one where you simply *cannot* write horizontally without
severely crippling the system.) Even better, has anyone gotten vertical
writing to actually *work* in an application, say your browser?
The reason I ask is 'cos I was doodling with a prospective Tatari Faran
font some time ago, and actually made a Truetype font for it, only to
feel rather unhappy with it a little later, and junking the whole thing.
I decided instead to go with a completely vertical system. In this new
system, the glyphs stack on each other and resembles decorations of a
pillar (or a totem-pole-like thing), and there are left- and
right-diacritics. (Laevocritics? Gauchocritics? I tried looking for the
appropriate Greek root since -critic comes from Greek, but came up with
'aristocritic', which just sounds too lame.)
</ObConlang>
T
--
Life is too short to run proprietary software. -- Bdale Garbee
Replies