bug-ncurses
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: Program started by bash works, program started by rc.local shows pro


From: Hohl, Gerrit
Subject: RE: Program started by bash works, program started by rc.local shows problems
Date: Thu, 15 Mar 2012 13:49:26 +0100

Hello everyone,

in the meantime I download the new ncurses 5.9 package from the Ubuntu 11.10 
branch and installed it in my 10.04.4:
libncurses5_5.9-1ubuntu5_i386.deb
libtinfo5_5.9-1ubuntu5_i386.deb

My program works the same as before: No problem in the bash shell, but still 
problems if it is started from the /etc/rc.local file.

Means that problem is not limited to ncurses 5.7, but still exists in ncurses 
5.9.

Regards,
Gerrit

-----Ursprüngliche Nachricht-----
Von: Hohl, Gerrit 
Gesendet: Donnerstag, 15. März 2012 12:44
An: 'address@hidden'
Betreff: Program started by bash works, program started by rc.local shows 
problems

Hello everyone,

I'm using Java and the Java Curses Library ( 
http://sourceforge.net/projects/javacurses/ ) which is some kind of wrapper for 
the ncurses library. If I start my program from the bash, I don't have any 
problems. As my program is some kind of setup program, I want to start it from 
the /etc/rc.local on an Ubuntu 10.04.4 LTS 32-bit system (libncurses5 
5.7+20090803-2ubuntu3 depending on lic6 >= 2.4). But if I do that the program 
shows some odd problems: The "rendering" of the screen is not correct. The 
shadow of my windows is on the left border on the screen instead of in the 
column before the last column. And a few other things are also upside-down.

I already opened a bug at the Java Curses Library website:
https://sourceforge.net/tracker/?func=detail&aid=3504280&group_id=37998&atid=420913
There I also added a screenshot which shows the problem.

As I didn't know if the problem was caused by that library or the ncurses 
library, I did some debugging: I created a log containing every move() and 
addch() (means the y and x coordinates as well as the character which was 
drawn). I compared the logs from the program startet from the bash and the 
program started by the rc.local. They were identical - in size and content. 
Means the Java Curses Library does the absolutely identical ncurses calls in 
both cases. So I guess it is a bug or problem within the ncurses or one of the 
libraries that are used by it.

Did someone have similar trouble in the past? Is there any bugfix or workaround 
available?

Regards,
Gerrit



reply via email to

[Prev in Thread] Current Thread [Next in Thread]