Descriptive Problem Summary:
When scaling an atom's transform under 512 there's a discrepancy where it translates downwards the further upscaled it becomes. I haven't tested if this relates to all matrices.
Numbered Steps to Reproduce Problem:
Ran under 512.
Ran under 511.
Test for yourself by running http://www.byond.com/developer/Kidpaddle45/MouseWheelZoom
Dec 10 2017, 10:07 am