yes, except its up a layer. remote desktop moves pixels, and X is a protocol that distributes the rendering api. so you start up your app, and query what kind

yes, except its up a layer. remote desktop moves pixels, and X is a protocol that distributes the rendering api. so you start up your app, and query what kind of display you're connected to - there is all kind of cruft about color models and bitplane depths. choose fonts, register for events (mouse, resize, etc), draw your elements and wait for the user to interact with you. it makes absolute sense in the context of the early 90s 'lets make this a network api' - and almost none from a modern perspective as bloated as the browser model is, i think shipping code to run on the users machine is a much more powerful model... if you can keep it small and deal with the security issues.