This project was a “scratch an itch” sort of project. I always found completion in Emacs to be annoying for one reason. When I perform a completion, the location of the point (Emacs parlance for cursor) on the screen often times changes. This forces me to go searching for where the point has moved to in order to see the result of the completion. Basically, when using Emacs, I suffered from these problems. It seemed that there was no reason for the point to move at all as there is plenty of screen real estate for the completions pop-up window. So I created this hack in order to keep the location of the point on screen fixed during completions.
I call this a hack because it doesn’t really use the normal Emacs channels to
work. There are no customizable variables or major/minor modes. There is a
function to enable the behavior and one to disable it. There is also a
variable, *completion-vertical-popup-size*, that determines the number of rows
the pop-up window should have. And that’s about it.
To use Fixed Point Completions, just download the file and place it in your load
path and use M-x load-library RET fixed-point-completions RET. To enable the
completion behavior, run the function (enable-fixed-point-completions). If
you want to do this on startup, you can just add this to your .emacs file:
(load "/path/to/fixed-point-completions.el")
(enable-fixed-point-completions)And, naturally, running (disable-fixed-point-completions) will return Emacs to
it’s normal completion behavior. While Fixed Point Completions is enabled, is
should alter the way all completions happen in all of Emacs.