demand for better unicode support in 21.4?
Julian Bradfield
been enhancing Unicode support for the 21.4 ("stable") codebase.
I'm wondering whether anybody else cares enough that I should release
something.
There are two projects:
(1) A fully native Unicode xemacs; this works for me now, but is a very
long way off being complete enough for what I would call release
quality.
(2) A set of "minimal changes" to 21.4.22 that allow the appallingly
unrobust mule-ucs code to be junked, and unicode to be handled by
simpler means. This works fine for me, but is missing some
necessary features (e.g. utf-16). However, it would only be a few
days' work to add those features. If other people are sufficiently
attached to 21.4 that they might use it, I could do that work.
Drop me a line if you're interested.
Yves Forkl
> As those on xemacs-beta know, I have, for whatever bizarre reasons,
> been enhancing Unicode support for the 21.4 ("stable") codebase.
> I'm wondering whether anybody else cares enough that I should release
> something.
Hello Julian,
great to hear that you have been investing quite some time into
significantly improved Unicode support in the current stable XEmacs
version. I, for one, do care very much about the work you have done -
it might prove to be the single thing that could keep me from having
to switch, after 20 years, from XEmacs to GNU Emacs, just because I
need at least minimal Unicode support. [1]
> There are two projects:
> (1) A fully native Unicode xemacs; this works for me now, but is a very
> long way off being complete enough for what I would call release
> quality.
> (2) A set of "minimal changes" to 21.4.22 that allow the appallingly
> unrobust mule-ucs code to be junked, and unicode to be handled by
> simpler means. This works fine for me, but is missing some
> necessary features (e.g. utf-16). However, it would only be a few
> days' work to add those features. If other people are sufficiently
> attached to 21.4 that they might use it, I could do that work.
While both projects certainly are equally worth pursuing, in the short
term I would prefer a smaller range of reliable functionality to a
complex concept from which only some loosely connected fragments are
actually implemented. So for the moment, I would happily accept having
to keep to editing UTF-8 files only if this means I can now reliably
edit Unicode-encoded files at all. (Conversion between Unicode
encoding schemes can be done using external tools.)
I am looking forward to whatever you want to release that will help
XEmacs 21.4 catch up in terms of Unicode support.
Yves
Footnotes:
[1] With "minimal Unicode support" I intend to mean the ability to
open, modify and save any UTF-8 encoded file without any data loss.
Currently, XEmacs 21.4.21 destroys numerous characters even when set
up correctly to deal with UTF-8 files, and unfortunately, German
quotation marks are among its victims.
Michael Piotrowski
On 2009-12-05, Julian Bradfield <j...@inf.ed.ac.uk> wrote:
> As those on xemacs-beta know, I have, for whatever bizarre reasons,
> been enhancing Unicode support for the 21.4 ("stable") codebase.
> I'm wondering whether anybody else cares enough that I should release
> something.
Yes! I'd be very happy to see better Unicode support in XEmacs 21.4.
> There are two projects:
> (1) A fully native Unicode xemacs; this works for me now, but is a very
> long way off being complete enough for what I would call release
> quality.
> (2) A set of "minimal changes" to 21.4.22 that allow the appallingly
> unrobust mule-ucs code to be junked, and unicode to be handled by
> simpler means. This works fine for me, but is missing some
> necessary features (e.g. utf-16). However, it would only be a few
> days' work to add those features. If other people are sufficiently
> attached to 21.4 that they might use it, I could do that work.
I can't say off-hand what would better fit my needs--both sound
interesting. Personally, I don't really need UTF-16 support but since I
work with medieval texts, which contain stuff such as LATIN CAPITAL
LETTER A + COMBINING LATIN SMALL LETTER E, LATIN SMALL LETTER U +
COBMINING LATIN SMALL LETTER O, or LATIN SMALL LETTER W + COMBINING
VERTICAL LINE ABOVE, *some* support for combining characters would be
great.
Displaying combining characters correctly is probably hard, but I'd
already be happy if they were preserved--currently it's not possible to
edit files containing combining characters at all since the combining
characters are replaced by question marks, which effectively destroys
the text when it's saved...
--
Dr.-Ing. Michael Piotrowski, M.A. <m...@dynalabs.de>
Public key at <http://www.dynalabs.de/mxp/pubkey.txt> (ID 0x1614A044)
Andreas Röhler
Hi Julian,
think at sxemac...@sxemacs.org that might be of some interest too.
Thanks
Andreas