Does a window need a graphical environment to exist? - linux

This one is a pretty strange question of mine so I'll try to explain the best I can.
On a DOS Linux distribution, without having installed any graphical environment, if I start an application which is supposed to have a GUI, will I be able to interact with that window via code?
What I thought is that if KDE/GNOME's role is to graphically represent an existing GUI, I should be able to interact with it but if these graphical environments have an active role in the window creation there is nothing I can do.

You won't be able to run a graphical program without a window manager.
The reason is that the window manager is doing a lot more than just allowing the window to be shown to the user. It comes with libraries and services for constructing windows and components and interacting with them. Without these libraries and services, the program won't be able to start.
The only exception would be if a program contained all the graphical code built into it, and didn't rely on a window manager. Realistically this would only happen for a fairly simple program and in special cases; for instance, Raspbian includes OMXPlayer, which shows full screen video without relying on any window manager.

Quoting (with some adaptation) from https://raspberrypi.stackexchange.com/a/3974 which answers how to start mplayer in X11 without a window manager:
The X server is really usually called X and you can just start it. You can set the DISPLAY number as an argument along with some other things.
You might want to have some sort of session, though, and still go through xinit or such and start X with startx. You can use .xinitrc or such as a script and simply not start a window manager there. You will need a "magic client" as the last command that stays running so X doesn't terminate immediately. In a traditional "failsafe" session, that was an xterm. If you know what to launch and launch once only, it could be that program or any UI/wrapper that does the launching for you. When the last ("magic") client in the init script terminates, X terminates.
If all you want to do is play video, you might see if you can run mplayer with directfb instead and skip using X11 entirely.

Related

Crate that allows my application to launch a new console screen

I have a Rust application that runs from a console. When run I want the application to run a thread that opens a 'new console' window and prints text to that window.
Is there a crate or configuration anyone can suggest for launching a second (and maybe third and fourth) Terminal screen?
That's very platform specific. You could start a new process, use x on Linux in process, use Wayland on Linux in process, use Win32 API on Windows, (list goes on)
You also might consider any ui frameworks etc for rust (of which there are few)
Creating a new process of terminal emulator on the system and writing to it's std in while launching cat in it is probably your best bet https://doc.rust-lang.org/std/process/struct.Command.html
I hope this helps

Hardware+OS setup for an Application

I want to make a system which would get input from USB barcode scanner, validate it on remote server and display an answer (text and images).
I would use JavaFX or in-browser JS web application to grab scanner input somehow.
I'm planning to run this application on Raspberry Pi or plain PC.
Is there a way (or a special linux distribution) to ensure that system always loads the same way and starts JavaFX app or opens particular web page in browser?
e.g. no login page, NO update or other popups are shown.
Any other ideas? Thanks
...
Found an interesting solution here, where JavaFX GUI app is started from command line, without X-Server(linux graphical interface) at all.
I have recently done something similar, and it's not too hard. Obviously working on the Pi, you will want to use something like Linux and having limited hardware will mean that you should be minimising what you are running. Due to this, I would recomend that you run a light weight distro. Something like arch (which is what I used) allows you to build only what you want from the ground up without the need to find and compile everything like you do for LFS or Gentoo.
As for booting, the following two wikis will give you the details of starting the Gui without manual login:
https://wiki.archlinux.org/index.php/Automatic_login_to_virtual_console
https://wiki.archlinux.org/index.php/Start_X_at_Login
After that, it is simply a matter of putting the command to launch your program in .xinitrc.

How does xTerm communicate with the XWindows session that contains it?

I want to script opening chromium in kiosk mode under Linux (openSUSE 12.3) and XWindows, while controlling which monitor chromium appear in. I also need to have this script be run outside of a standard xterm, e.g. via ssh into the machine where chromium AND the display will run (i.e. I want the application to appear on one of the displays of the machine I'm connecting to, not the machine I'm connecting from).
I can do everything I want by hand in xterm on the target machine, but I won't be running these scripts in xterm. Somehow, the xterm environment is special: it knows how to talk to the containing XWindows session, and in fact knows which actual monitor it is hosted on. For example:
When I ssh into the machine and try to run an XWindows GUI on the target machine's display, it (reasonably) complains "Can't open display". However, even when I set the DISPLAY environment variable to the same value I see in xterm, I still get the error (although it now includes the current DISPLAY value). I've also tried setting DISPLAY to a variety of permutations on ":X.Y", where X and Y are small integers, but again to no avail.
If I run chromium in --kiosk mode from xterm, it takes over the monitor that currently contains the xterm window. If I drag xterm to my other monitor then chromium will launch in that monitor. I want that control when I launch chromium from my script, but can't figure out the background magic.
For window manipulation, I've seen suggestions to use wmctrl or devilspie, but I'm a bit nervous about depending on tools that haven't been updated in years. I'd also prefer to have chromium launch in the right place right off the bat, rather than having it appear in the wrong place and then be moved. And neither app can help me launch the app remotely.
So, here are the questions. First, how does xterm tell chromium how to connect to the right XWindows session and monitor? Second, how do I simulate that from a completely separate ssh connection? (Again: it's more than just the DISPLAY variable.)
Thanks,
Dan
You can't open X client on the target machine's display because of permissions: not anyone that can ssh into a machine may "take ownership" of it's screen. These are things controlled by xset -- read up on it.
As for the display: I assume that the specific monitor is controlled by the second part of the DISPLAY environment. At least that's what I remember from reading about X. I've never actually worked in a multi-monitor X environment.
Xterm is not special. Your mouse pointer is. Chromium will appear on the monitor your mouse pointer happens to be in. This is probably the default behaviour of your window manager.
Chromium does not respect the usual window position requests. You cannot tell Chromium where to appear, you must tell your window manager where to place the Chromium window. Methods of doing that, if they exist, totally depend on what window manager you have.
If you ssh to your machine as the user that has started the X session, you should have no problem connecting to that session.

Communicating with another terminal application via SSH

I'll start off by saying I'm no Linux guru, or even close. I use it to develop embedded applications and it serves my purposes just fine as such.
I have a program which is running (I suppose the correct terminology is "running in a terminal") on a CentOS 6.3 box. For debug and statistics, I have a routine which monitors the keyboard (stdin file) and spits out items as requested by the given key-presses. That all works fine when I'm sitting in front of it.
But I'd like to be able to perform these simple functions: press a key, see some output, remotely. I can SSH into the box and execute commands, but I can't "see" this program.
I have tried searching "communicating with other terminals using ssh" and myriad variations, but I guess I'm not asking this correctly -- the search results are worthless.
What I'd like to be able to do is login to my account and then somehow "see" my program running and type keys and see its output. Is this a stdin/stdout redirection issue?
I apologize in advance if this is painfully obvious and I'm just an idiot, but I'd still like to know how to do it... :)
If you use the linux 'screen' utility, you can re-attach to the original terminal session that you used to start the program. basically you just type screen and then run your program. here is more useful info on screen:
http://www.rackaid.com/resources/linux-screen-tutorial-and-how-to

How to start a gtk application after booting up without logging in?

I have got a gtk application. I want it to start automatically when I boot up the system, such that instead of getting usual login screen, my application greets the user.
My application is a sort of cash dispensing application. So user should not even see the login screen of linux. Whenever machine in boot up, after loading services and all that, my application should appear on the screen. One should not have to put any password or login in anyway to start that application.
Now so far I have tried the following but all in vain.
I put the command in my /etc/rc.local file
/home/EXE/cashier
Then rebooted machine in run level 5. But nothing happens. I get the usual login screen. Other programs that do not involve any gui or gtk, they are automatically started in this manner but not the gtk application.
Then I put
xint /home/EXE/cahier
and later
xint /home/EXE/./cashier
But they too didn't work. When I looked into details, I found out that "An Xserver is already running on screen 0, cannot open other server" or something like that.
Then I tried booting into run level 3, this time it worked.
xint /home/EXE/.cachier
But in run level 3, the font and icon sizes are smaller. Moreover some windows do not cover the whole screen.
Now I will really appreciate a way to run my application in run level 5, without showing login screen automatically.
Regards
Edit
Currently, the application is running on fedora core 2
This is more a question about linux distribution startup than it is about Gtk applications. The specifics are going to depend on your distribution, but the essence is:
Disable the gdm facility, which is normally responsible for spawning your X server.
Run xinit (not "xint") to start the server manually. This takes a client as its argument. It may work to run your Gtk application directly (it needs to be smart enough to position and size itself instead of relying on a window manager), but more likely you will want to pick a window manager to run it in. The easiest way to do this is via script that you pass to xinit.

Resources