It’s even more overloaded when you include client/server as a paradigm in application development. Sure, the server is still a program usually listening on a socket, dbus name or using shared memory of some sort an and responding to requests but they’re also logically part of the same program.<p>tmux, evolution, mutter, urvxt, tracker, emacs, neovim, etc. are all servers in the right context.<p>Then there are servers that don’t serve requests in the traditional sense. Entities that perform batch operations by pulling data from queues, run reports, etc. are also called servers despite not fitting the call & response model.