SOS! Same old story (w32 internationalization problem)

SOS! Same old story (w32 internationalization problem)

[Edward Casey]

I have now gotten gnu emacs to work on 2 desktops, one win2k and one
win98. On the laptop (Compaq Armada E500) I have not been able to get
internationalization to work correctly no matter what I do although all
the other functions of the editor seem to work. A very knowledgeable
poster from this newsgroup assured me that the problem had to be related
to the unavailability of the correct fonts on the laptop. I rebooted the
laptop in dos, ran deltree on the windows\fonts subdirectory and then
xcopied the exact same fonts that are on the win98 desktop into a new
windows\fonts subdirectory. After rebooting in windows and starting
emacs, if I type C-x RET c utf-8 and then open a known good file saved
as utf-8, I see all the correct glyphs on the laptop but I can only add
new ones by copy-pasting, but not using leim latin-4-postfix or a
macronize function defined in the .emacs.

 From *Messages* buffer:

(C:\EMACS\EMACS-21.3\BIN\EMACS.EXE --debug init)
For information about the GNU Project and its goals, type C-h C-p.
Loading quail...done
Loading quail/latin-post...done
Auto-saving...done
Loading mule-util...done

The following 5 characters were produced with latin-4-postfix input
method (vowel + dash) but they appear as empty rectangles. They were
identified with key combo C-x =

Char: ā (05140, 2656, 0xa60, file ...) point=18 of 30203 (0%) column 0
Char: ē (05072, 2618, 0xa3a, file ...) point=19 of 30203 (0%) column 1
Char: ī (05157, 2671, 0xa6f, file ...) point=20 of 30203 (0%) column 2
Char: ō (05162, 2674, 0xa72, file ...) point=21 of 30203 (0%) column 3
Char: ū (05176, 2686, 0xa7e, file ...) point=22 of 30203 (0%) column 4
keyboard-quit: Quit
exchange-point-and-mark: No mark set in this buffer

[The following 5 characters were in the same file and appeared
correctly. Also identified with C-x =

Char: ā (01210041, 331809, 0x51021, file ...) point=226 of 30203 (1%)
column 189
Char: ē (01210063, 331827, 0x51033, file ...) point=157 of 30203 (1%)
column 120
Char: ī (01210113, 331851, 0x5104b, file ...) point=286 of 30203 (1%)
column 11
Char: ō (01210155, 331885, 0x5106d, file ...) point=316 of 30203 (1%)
column 41
Char: ū (01210253, 331947, 0x510ab, file ...) point=618 of 30203 (2%)
column 189
Mark set [2 times]

Are there any w32 gnu emacs experts on internationalization out there
who might be able to help me with this?

Thanks,

Ed

Re: SOS! Same old story (w32 internationalization problem)

[Jason Rumney]

"Edward Casey" <ejmn@cpinternet.com> writes:

> After rebooting in windows and starting emacs, if I type C-x RET c
> utf-8 and then open a known good file saved as utf-8, I see all the
> correct glyphs on the laptop but I can only add new ones by
> copy-pasting, but not using leim latin-4-postfix or a macronize
> function defined in the .emacs.

How about defining the macronize function to insert utf-8 versions of
the character instead of Latin-4. It seems that your PC can display
Unicode, but cannot display Latin-4 (it is probably missing the
Latin-4 codepage, which I think is defined in the file CP_28594.NLS)

Re: SOS! Same old story (w32 internationalization problem)

[Edward Casey]

<jasonr (Jason Rumney) @ f2s.com> wrote in message
news:uy8heze87.fsf@jasonrumney.net...
> "Edward Casey" <ejmn@cpinternet.com> writes:
>
> > After rebooting in windows and starting emacs, if I type C-x RET c
> > utf-8 and then open a known good file saved as utf-8, I see all the
> > correct glyphs on the laptop but I can only add new ones by
> > copy-pasting, but not using leim latin-4-postfix or a macronize
> > function defined in the .emacs.
>
> How about defining the macronize function to insert utf-8 versions of
> the character instead of Latin-4. It seems that your PC can display
> Unicode, but cannot display Latin-4 (it is probably missing the
> Latin-4 codepage, which I think is defined in the file CP_28594.NLS)

Thanks a million Jason.  That and half a dozen other codepages were
missing from the laptop.  I decided to keep the setup the same on all
three machines. After copying the *.nls files over all I had to do was
add entries to the registry with regedit. I had thought that the
codepages loaded with Baltic language support but maybe they are with
Turkish instead.

Ed.