Let's say that I have a customized Bash prompt stored as a PS1 variable. My variable gets read every time I invoke a login shell, i.e., when logging in in one of the standard virtual terminal or by invoking xterm with the -ls parameter.
However, when I just invoke the terminal without specifying that it is a login shell, I just get something like: bash-XXX$ as a prompt.
where should I place my PS1 variable so that it could get read even on non-login shells?
I am having trouble distinguishing between login and interactive shells as they relate to terminals. I understand that a login shell is what is started when you login, and an interactive shell is used by scripts and such. I also get that terminals will use an interactive shell by default. What I don't understand is why interactive shells have no prompt by default. Is their a way to get an interactive shell to inherit the normal PS1 prompt? Is it considered bad to do this? Would it be better to have terminals start login shells, and if so why or why not?
For example, if I'm in csh, I can use `setenv VARNAME varVALUE` while I can use export in Bash. Given that the environmental variables are created, can BASH read env vars from csh and vice versa?
I am trying to figure out where variables are set for interactive shells? In particular, I am trying to have LS_OPTIONS inherited by interactive shells as it is by login shells. I understand LS_OPTIONS is set in /etc/profile, and this may not be processed by interactive shells but by login shells. However, I also note other variables such as PATH and INPUTRC are set correctly in interactive shells, and these are also set in /etc/profile from what I understand. So how is it determined which variables are inherited by all shells, and which are just for login shells?
I want to have an ls' output colorization in gnu screen. Colorization in my system (Slackware 13) is realized by aliasing of ls in /etc/profile.d/coreutils-dircolors.sh:
Code: $ alias ls alias ls='/bin/ls $LS_OPTIONS' where $LS_OPTIONS is Code: $ echo $LS_OPTIONS -F -T 0 --color=auto
But in screen this alias isn't defined. It seems like /etc/profile script isn't executed at shell starting in screen. I think it happens because screen starts a shell not as a login shell. I tried to correct it by adding to ~/.screenrc or to /etc/screenrc. The problem is the same. By the way when I start screen as a root I haven't this problem. What's wrong?
Quote: The precompiled Slackware kernels are available in the /kernels directory on the Slackware CD-ROM or on the FTP site in the main Slackware directory. I am unable to reach it, what's the proper login?
I finally decided to take the plunge and try out slackware64 on my new laptop, and making the switch has been surprisingly painless. One issue I still haven't managed to sort out is getting hal to read /etc/hal/fdi.I have two files in /etc/hal/fdi/policy: one is a copy of 11-x11-synaptics.fdi, and the other is a series of rules to set up the trackpoint on my thinkpad. Both of these files were read and added to hal when I was running 32-bit 13.0, but they don't seem to be read in 64-bit. I've looked at lshal, and I don't notice anything glaringly wrong there, but I'm not sure what else to check.
Does anyone know of any free services offering debian or ubuntu shells? I want to run IRC remotely with screen, but free services only seem to offer unix shells. I'm not really bothered by that but I feel more comfortable on ubuntu.
I am newer to Slackware but not to Linux. I am having the hardest time trying to get my CDROM to mount for me so that I can play audio cd;s and the like. I can see that the system knows that my CDROM drive is there and that it works because I installed Slackware 13 from my CDROM drive as /dev/hdc. Now when I put in a cd nothing happens? I have googled away and searched this forum quite a bit before posting this but I have tried almost everything I can think of and what others have mentioned. I have screenshots to show you what I mean. When I try to run the command: mount /dev/cdrom /mnt/cdrom I get it as mounting as read only and it says can't read superblock? Now that is fro my slave IDE CDROM on top as hdd. I have tried before to mount it as /dev/hdc but to no avail. I know both CD drives work just fine because I used them to install the system. I am going to put the screenshots below so maybe someone can help me! Also I can eject the top CD drive with the command code...
Yesterday, I bought a 1TB WD Passport, for backup and storage. It uses NTFS, and I've had no problems manually mounting and moving files to and from it from root. However, I don't like having to be root to in any way modify the data on the drive. In order to avoid this I decided to create a line in fstab that would allow permissions to the user, so I added this to my fstab:
This allows me to mount, unmount, and peruse the external HDD - however, if this is active, neither the user NOR root have permission to make any changes. The HDD acts as read only, even though there is no "ro" option on my fstab.
While I was installing some packages, the power in my street went out (uncleanly killing my system) ... unfortunately Kpackage was open at the time of the power outage. Now when I try to use KPackage, I get "Login Problem, Please Login Manually" ... which does not allow me to login using either the root or user's password. I have rebooted my computer many times and have even run a 'reiserfsck /dev/*** --fix-fixable' on my system, which did not repair the problem.
I have researched this error, but have only found responses that people suggest changing the KPackage behaviour from 'su' to 'sudo' ... ; while this does work, it feels like a 'putting-a-band-aid-over-a-warning-light-so-I-can't-see-the-warning-light-anymore' kind of 'fix?' which isn't good as it would enable anyone using the system to add or remove packages without thought or consequence. What I would prefer to do is actually fix the problem so that proper root password entering is again required to add or remove packages, so my question is:
1. Does anyone know if KPackage 'locks' out a file(s) on the system which may be preventing me from logging into the program correctly, and if so what and where the file(s) may be?
2. Can I delete any kpackage (profile?) file to gain normal login behaviour again?
3. Is this a permissions error?, has something in users/groups? been broken that I can look into?
4. how I may be able to properly repair this KPackage login problem on my system?
I'm a new Ubuntu user and a Python programmer, it's the first time that I use Python in Ubuntu so it's a bit confusing me. If I want to save a Module or something in a specified map, I get 'Errno 13', it says that I don't have permission to edit, do thing in that map. And this is also for importing files with Python. I logged in from Terminal with 'sudo -i' and closed Terminal, but the problem keeps repeating. How do I login as Admin or Root and stay as Admin or Root? I need right to edit/read files as well as root and normal user.
An example: Python file: test.py -> test = open('/a.txt', 'w') test.write('Test - Test - Test') test.close
When I execute this, I get in the Python IDLE the 'Errno 13' problem and below the 'Errno 13' it says that I don't have permission. Who know how to stay logged in as root, even if the user has not logged in or isn't this possible, if it isn't then I just want to get files moving, editing/erasing etc.
I was able to set up SSH and log into my linux machine and I want to interact with shells already running on the desktop. I can locate and view the processes but I'm wondering if its possible to "get into" the program I already have running via SSH and enter commands.
Is there any way to have x server on multiple shells at a time?(eg. Sally is logged in on shell 6 with her own desktop cube while Rob is logged in on shell 7 with his own desktop cube, etc)
I'm finalizing the the dual-boot addition that I recently completed. I've added lubuntu to the LT3114u Netbook that I use. That's done, now I'm migrating my user environment back from the Cygwin one that I've been using on Windoze, as of late. One of features that's been useful has been the global environment management possible with the windoze UI functions. As I coded my emacs environment while 'over there' (windoze side), I included a check for the lisp library repository path environment variable EMACSLOADPATH. It is globally available when one logs in, and does not get set in any user scripts, or dot files any to do with CYGWIN or emacs.
While it has been easy enough to set EMACSLOADPATH for the login shell environment, via .bashrc, .profile, and system wide via /etc/profile, I have not been able to have '/usr/bin/emacs23' have EMACSLOADPATH set when invoked via the LXDE UI. Thus, it would appear that emacs is not being initialized with the same environment that any of the shells are (old school?). Thus, what is it that one must update in order to have GNU emacs initialized with the same environment variables as bash? Or, put another way, why aren't the environment variable defined in /etc/profile, ~/.bashrc and/or ~/.profile ending up in the initialized environment upon invocation of emacs23 on lubuntu Linux?
I have a problem when i ried run this command (export HISTTIMEFORMAT="%F %T ") on slackware 13.0. the answer is the following (-bash: HISTTIMEFORMAT: readonly variable) I am running this command on the bash like root.
When I try to go to a virtual terminal from X the writing is too dim to read. When I boot to run level 3 all is fine, then I "startx" and use the gui. If I "ctrl-alt-F3" the printing on the screen is so dim that I can't see it. Is there a setting for this somewhere that I haven't found. Before installing Slackware 13-64 I ran Zenwalk for over a year and never had this problem.
The author (somewhere on these Ubuntu forums directed........ Ctrl + Alt + F1 which I thought was going to be the "Terminal" but nothing similar. I have about a half dozen different names/passwords for Ubuntu but none will work to log in. How to get out of this text console and back to Ubuntu GUI? And, how to read my lost LogIn name and password for this "Text Console"?
How to configure the KDE Konsole that when I press on a "new tab" button, the new shell would open on another machine (thorugh ssh), and not on the current one?
I changed shells for a user from bash to tcsh using this command... chsh user -s /bin/tcsh but when logging on via ssh under that user I get a "Illegal variable name" error. I type, echo $shell and it says, /bin/tcsh. I am curious what that Illegal variable name could be.
I am using Linux some years, but since I built a LFS, I feel noobish again. Now with the help of BLFS I am setting up my environment and somewhere I incidentally read, that not every variable is inherited by a child Bash shell. As for the $PS1 and $PS2 variables I know, that they are not inherited by non-interactive Bash shells (and there is no reason why they should in my opinion). Well, as for my first thread I hope the title gives enough information on what I want to know. But anyways: Which environment variables are not inherited by Bash shells?
I'm getting something(s) wrong, trying to run commands (both simple and piped) in shells from Perl programs. The ultimate objective is to set up "copy X selection to clipboard" from urxvt but apparently simple debugging statements are not working.Here's the Perl, taken from here and modified to use xclip instead of xsel and with debugging added, shown in green
