All OK. I just installed elektra from OBS (from SLES 9) instead of the one I found in Fedora Core 10 and now all works great.
This was installed for Cinepaint.
All OK. I just installed elektra from OBS (from SLES 9) instead of the one I found in Fedora Core 10 and now all works great.
This was installed for Cinepaint.
I think that when you look inside that file you will find things like “command not found” because that is what your original error messages complain about.
To me it looks that during an installation a script was executed to create that file, but the script had errors and the programmer did not seperate standard output and error output. Thus the error messages went into the file.
Also I think that the original script (which made those errors) shows how dangerous it is to install packages that are only tested on another distro (or may be even only tested on the one installation the packager had). It calls tools that are not on another distro and there you go!
I fully support the others that you have to deinstall everything that may remotely has to do with this. Who knows what other havoc is done? Also then you are not sure that everything is OK. When somebody packages like this, who trusts his deinstallation procedure? Not me!
Ansus wrote:
> It was /etc/profile.d/elektra-elektraenv.sh
maybe you can un-run it, or run it backwards…
–
palladium
I simply can delete it and all works well. But I decided to change the package elektra from that of FC10 to that from SLES9 and now there is no error reported.
Also I think that the original script (which made those errors) shows how dangerous it is to install packages that are only tested on another distro (or may be even only tested on the one installation the packager had). It calls tools that are not on another distro and there you go!
Yes! You see, it even created an environment variable named “Command” without any value: obviously a result of that a part of error message “Command not found” became argument of ‘export’ command.
Did you learn nothing from the previous posts and experience? SLES9 is NOT openSuSE!!! They may both be from Novell but they are not the same. Different libraries and environments.
tommyttt wrote:
> Did you learn nothing from the previous posts and experience? SLES9 is
> NOT openSuSE!!! They may both be from Novell but they are not the same.
> Different libraries and environments.
far as i can see:
information is flowing at the speed of heat…
but, learning is occurring at an imperceptible glacial pace (if at all)…
its kinda like:
if
SLES9 is Linux
and
Fedora Core 10 is Linux
and
openSUSE 11.2 is Linux
then
SLES9 or Fedora Core 10 is openSUSE 11.2
good luck with that, Ansus…
–
palladium
Just as a final comment:
(It does of course NOT ensure that everything is ok!)
Run at least “zypper verify” to see if there are any obvious conflicts in
your system.
Здравствуйте Ansus!
Why do you want to use su -, if su without - works fine?
su and su - produce different environments. su alone is a mix of the user and root. su - is more pure root.
gogalthorp wrote:
> su and su - produce different environments. su alone is a mix of the
> user and root. su - is more pure root.
+.75
if you had said
su - is pure root. (or pure root only)
i would have given you +1
but, if you had added: http://tinyurl.com/ydbwssh
you would’a had +2 !! get those score cards ready…
–
palladium
By the way I wonder why sudo cannot start graphical applications?
Sudo does not start applications a all.
What you mean is that when you start an application using sudo to change to another user, the application can not open a window on your xserver. That is correct. Imagin that everybody that is running on your system (never forget Linux is a muliti user system) or that can reach your system (potentialy everybody on the Internet) could open windows on your desktop!
Thus the one who runs the xserver (in this case you as the user who did the sudo) must allow this. The statement which does this is xhost. So read;
man xhost
and find out why
xhost +local:0.0
is doing the trick. More in several other threads here on the Forums.
hcvv wrote:
> --------------------
> man xhost
> --------------------
i don’t understand the answer to your ‘why?’…but, i never question
Henk as he knows more than i do…
but for your why question (Ansus) i give you the easy (but still right
enough way) answer to the how question:
wanna open kate as root while running KDE, then use this:
kdesu kate
or, wanna run a root powered nautilus in gnome:
gnomesu nautilus
of course if you have both kde and gnome you can also do:
gnomesu konqueror [and etc]
WHY do kdesu or gnomesu? because it solves the problem described by hcvv
btw, this will also works:
su - root -c kate
–
palladium
As always I answer to the question as posted: Why?
And palladium answers to the question as (may be) intended: How?
And thus we have the best of both worlds lol!
That is correct. Imagin that everybody that is running on your system (never forget Linux is a muliti user system) or that can reach your system (potentialy everybody on the Internet) could open windows on your desktop!
Well to do so one should have a root password. And under Ubuntu it works out of the box.
xhost +local:0.0
Some people said me that this creates a vulnerability in the system. I dont know, why, but in order to run PlaneShift or Alpha Cerntauri or (reportedly) Quake3 you have to run xhost +localhost before the xorg starts. I filed a bugreport since many games do not start because of this issue: https://bugzilla.novell.com/show_bug.cgi?id=586096
but it was closed, I wonder, why? Maybe I should better formulate it?
You did not quite get the message. It has nothing to do with root. When I say from the Internet, I mean from the Internet. There is no root user on the Internet.
I can start any X-windows appllication on any x-server (that is why it is a server) that I can reach via the net:
DISPLAY=computer.at.domain:0.0 xclock
would start xclock on my system and display the window on display 0 on system computer.at.domain.
There can be no need to run an application like a game as root, never.
And when you searched the Forums about threads about xhost, you now will know how to see that it is always on in your GUI.
About the bug report. It is a bit short answer, but in principle they are correct. And they may have started from the assumption the one who started the bugreport read and understood the man page of xhost and the general principles of the X-server.
About security. When you have read and understood the man page, which I suppose you did because I refered you to it and you seem to have no questions about what you read, you will now also understand how much you compromise your security in using it for localhost only. And of course, when a distribution does this by default, you are compromised in the same way (without even knowing the details) because it is the same x-server with the same features.
Okay. What should I do to run the games under OpenSUSE withount introducing vulnerabilities and shouldn’t such games run on OpenSUSE out of box?
Games? Did I miss something? What is the problem?
Well to do so one should have a root password. And under Ubuntu it works out of the box.
Ubuntu has a… special way of using sudo and treating root, fortunately a way which is pretty much ubuntian only. You see, ‘sudo’ is actually not a command to become root, but to give a regular user root-privileges for single tasks. This is set in /etc/sudoers, which on SuSE (and almost any other system) is unconfigured, so no user may execute any command with root-privileges. Ubuntu on the other hand deactivates root fully and gives the first user the right to call any command with sudo.
This is set in /etc/sudoers, which on SuSE (and almost any other system) is unconfigured, so no user may execute any command with root-privileges.
It is easy to enable sudo on OpenSuse (for tasks like sudo zypper update etc). But sudo cannot run X applications unlike in Ubuntu.
Ubuntu on the other hand deactivates root fully.
Not correct. Root in Ubuntu exists just like in OpenSUSE. There is no difference.
It is easy to enable sudo on OpenSuse (for tasks like sudo zypper update etc).
Although this command does work, it is total nonsense when understanding the actual intention of ‘sudo’. Read ‘man sudo’ about it, you might recognize potential side effects too.
Not correct. Root in Ubuntu exists just like in OpenSUSE. There is no difference.
Um… would you accept Wikipedia to explain that?
While I don’t think this feature is giving any more security (just the opposite even), you can easily see Ubuntu takes a different road. Waaaay different.