Re: windows-1251 language environment

[Stephen J. Turnbull]

>>>>> "Stefan" == Stefan Monnier <address@hidden> writes:

    Stefan> How does this use of multiple language environments work
    Stefan> in practice?  I mean, the language environment setting is
    Stefan> global, so how does it interact with the various
    Stefan> pre-existing buffers when you switch?

I have a C-c binding to switch language environments, works like C-x
b.  Emacs doesn't know a buffer<->language environment mapping.

I suspect there may be underlying problems, for example with XIM,
because my usage is quite restricted and I just may not be seeing
them.  (I don't use XIM because my preferred IM is kinput2, which eats
meta characters in its Xt filter and thus conflicts badly with Emacs.)

    Stefan> What is changed between those environments other than the
    Stefan> preferred coding system ?

(1) The whole coding priority list.  On Unix I have lots of files in
ISO-2022-JP, and I need to bump the priority of little-endian UTF-16
too, so that I can (sometimes) get autorecognition of .doc files (I
don't know of a better doc2txt utility than Emacs for _Japanese_ .doc
at this time).  I don't do mail from Windows, and I have Word, so I
don't do those tweaks on Windows.

(2) If I were capable of handling other languages, the default input
method would change.  (This would probably have to tweak C-\'s notion
of current input method and its state of activation, too, my language
switch function does that when I switch to or from English.)

(3) If these hypothetical other languages were European, I'd probably
want my ispell dictionary to change, too.  (Yah, that's not part of
an Emacs "language environment", but maybe it should be?)

    Stefan> Is this arrangement natural, convenient, and intuitive,

Natural, well, we're talking about a perverse environment to start with.

I tend to work on one system, multiple tasks, for an extended amount
of time.  So I don't need to switch very often.  I think the "natural"
thing to do (in XEmacs) would be to make language environment be a
specifier variable (XEmacs's generalization of buffer/frame locals).

Actually, what I'd _really_ like is to automatically be in Japanese
language environment (with input method invoked) in regions of
japanese-jisx*, and in English environment elsewhere.  The whole POSIX
concept of locale (as a global variable that "gets set") gives me an
itch.  It makes sense in the POSIX environment, at least a little bit,
but Emacs should be able to do a lot better.

-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
               Ask not how you can "do" free software business;
              ask what your business can "do for" free software.