Re: [STUMP] UTF8 related bug

[Stefan Reichör]

Daimrod <address@hidden> writes:

> Stefan Reichör <address@hidden> writes:
>
>> Because I have learned today where to look up a crash in stumpwm, I have
>> another crash that happens from time to time.
>>
>> When I open a file containing an umlaut (e.g. ö) in evince, the whole
>> stumpwm session dies with the following backtrace:
>
> I put the following snippet at the top of my .xsession to set the LANG
>
> LANG="en_US.UTF-8"
> export LANG
>
> Though I've tried to reproduce this but I can't crash Stumpwm.
>
> I've tried in utf-8 stumpwm with iso_8859-1 filenames and in iso_8859-1
> with utf-8 filenames.
>
> Can you provide more details on how you open the file?

# echo $LANG
en_AU.UTF-8


Here is the content of /usr/share/xsessions/stumpwm.desktop:

[Desktop Entry]
Encoding=UTF-8
Type=XSession
Exec=/home/srei/prg/stumpwm
TryExec=stumpwm
Name=StumpWM
Comment=Stump window manager


The problem appears when I open a pdf file from a GNUS attachment that
contains an umlaut. So I guess it is rather specific to my setup.

I often have problems, launching files from emacs that contain umlauts.
It seems that emacs uses a different coding system for the file names
than the utf8 shell.

M-x getenv RET LANG RET also gives en_AU.UTF-8

When I look at a file in emacs that contains an umlaut it looks correct.
When I do ls in the shell, I see '?' instead.

The bad thing is, that opening such a wrong file kills my whole
X-Session. Before I thought it had something to do with a bad pdf
content. Now I know that the filename is the likely cause. So I can save
the file, rename it and then view it.

However, it would be nice to
a) prevent stumpwm from crashing
b) find a setup in emacs to handle filenames containing umlauts that works


Stefan.