…sues

…ng it down more and more)

…bunch of thrashing that needs to be fixed

…l haven't figured out how to stop the row height thrashing

…the game on close

…sure why this code still existed since if the size changes it was already being rebuilt

…t (~2%) but this is an extremely hot code path

…is empty... it saves so much time... from 1.065us/call to 0.215us/call, and this is a call that is getting hit around 37k times per second just on the main menu when locked to 30fps... unlimited FPS on my machine resulted in 357k hits per second...

at 30fps it was a difference between 39.4ms vs 8ms total, and uncapped it was 380ms vs 76.7ms

note: more performance can be saved here if we can figure out a way to track the emptiness of the thread queues in the tree without adding back that pesky parent BS in the TQ that destroyed performance

potential 3:45am braindead solution for this is to pass a reference to a counter or boolean in the node that the TQ has access to and this can be propagated up the tree or something -- that or maybe lazily creating the thread queue, or somehow having a flat global threadqueue lookup (though i think this has the potential to be non-performant in comparison)

…ints can be set properly

…elected resolution

… label it

…gging