- get back to full-featured state with current compiler
- check all commented lines to see what they were good for (//, /*, #if 0)
- squash all warnings
- reenable --benchmark
- and make it run from gitlab CI
- LSystems
- implement multithreaded transformations:
- add menu entry to switch transformation method
- make Appearance menu submenu of new Settings menu
- multithreaded transformations for Composites
- use Thrust library for CUDA transform
- add menu entry to switch transformation method
- fix crash when switching desktop and back ("... is not a GL list")
- make rotations on Composites behave correctly
- ensure commitDraw() is only called once, after all vertices are added - do
something with scope
- use a GL view that takes advantage of this optimization
- refactor test suites for all Displayables to use as much common code as possible
- Matrix should have late binding so different algorithms for multiplication can
be selected
- test how much of a performance hit polymorphic Matrix multiplication is
- move all creations of Matrix objects to a factory
- replace operator notation with explicit function calls ;-(
- make the output function of Matrix, Vector, Rotation, NestedVector and whatnot conform to same interface
- combine vertex array with Realm to have a single class for surfaces without interdependence
- continue with "torations" - or generalize rotations so they can take displacement as a parameter
- separate the UI::View functions out of C4DView and make the rest a controller
class
- change Displayable::Draw() to take a shared_ptr instead of a raw pointer