an installer automatically starts, or starts when a user clicks on the downloaded ‘package’.
the download automatically uncompresses itself (unzips) to a ‘package’ (a specialized archive of component files).
a user clicks a ‘Download’ button in a web page to start the process.
The modern, user-friendly sense of ‘install’ software And it often yields only a command line app. But that sometimes is a trip to ‘dependency hell’. That’s not necessarily hard: it is often reduced to three commands. Worse, ‘install’ could mean: download source code, compile it, and THEN install the binary. The name of the program file is the command to execute it. The user downloads a binary and fiddles with system directories to copy the executable to some ‘bin’ directory. The user uses a command line in a terminal or console. A user deeply understands the process of installing, and the OS. See also Strategies for packaging Python apps on a Linux platform The ancient sense of ‘install’ software This post is about getting your foot in the door: publishing an app outside the main channels of distribution, but that installs in a user-friendly way. (I intend to publish a followup with the specifics.)Ī similar tutorial about short-cutting Debian packaging, is about a command line app, and not user-friendly. Suppose you want to create an installer for the Linux platform for an app that you created using PyInstaller.