just tried. this idea is so cool.
while i didnt miss most of the result
i think make a new nip, which start calculation the time the mouse is moving on the screen, stop tracking when it release, publish an event(maybe kind 8086) about how much time user A spending on event B and fomostr will eat all kind 8086 events and generate the result.
just needs a better algorithm, this one I coded today very quickly and there's a loooot of room for improvement (like right now I'm not clustering your activity when determining the times you were active at)
also, the window size I'm using is very short because I wanted to find more events that would fit in the "inactive window", right now it's set at 30 seconds or something like that
Iām just joking šš«
