Most modern Unix-based operating systems (e.g. Linux, MacOS X, and BSD) come with a little console-mode utility called GNU Screen. It’s a powerful tool in the hands of the console warrior, a veritable Swiss Army knife of text-mode human-computer interaction.
This utility, despite its considerable usefulness, goes unused by a great many people. Why is this? Few people even know it’s there, and those that do rarely understand the problem it solves. In this article, I will explain what screen does for you, and provide a simple set of instructions for accomplishing basic tasks using screen. My audience is those that are skilled with the command line but who have little or no experience with screen itself.
What screen does
Screen is best described as a terminal multiplexer. Using it, you can run any number of console-based applications–interactive command shells, curses-based applications, text editors, etc.–within a single terminal. The desire to do this is what gets most people hooked on screen. I used to start a half-dozen terminal emulators when I logged into my machine: I wanted one to read my email, one to edit my code, one to compile my code, one for my newsreader, one for a shell into my web host, and so on. Now I start one terminal emulator, and run screen in it. Problem solved.
The other main cool feature of screen is its ability to decouple the terminal emulator from the running programs. This means that you can use screen to keep programs running after you accidentally close the terminal emulator, or even after you log out, and later resume right where you were. It means that the idea of a “session” in which you are running a number of console programs is a free-floating entity that you can bind to any terminal anywhere, or no terminal at all if you want.
Read more at www.kuro5hin.org