Moony
Moony
A single bug isn't worthy of mention, but ok. :P
I have a USB controller, and could possibly try it out. Only thing i can think of right now is just a virtual mouse controlled by the joystick, with A/B/X...
EDIT: This comment has been proven incorrect. Leaving it for context only. The bias is probably caused due to the fact the air sim doesn't try and separate old state...
@LBPHacker wait i thought i mentioned this before. I didn't? Oops. Well now it's an issue, so that helps. The +0.5f trick is *slightly* faster, so when negative numbers are...
I'll test in a little. I have a offbrand(!) XBox controller with rather silly precision on the joystick, so it'll make good testing.
'changes'? You mean a rewrite of the font rendering? :P
Looking through aircode, it might be a good safety measure to replace the various instances of (int)(anumberbeingrounded + 0.5f) with std::round if the number cannot be guaranteed positive. The issue...
GCC provides a trival way to handle getting an error trace for errors. execinfo.h For MSVC (no GCC), [StackWalker](https://github.com/JochenKalmbach/StackWalker) would be a dependency if we don't roll our own MSVC...
Would it be possible to just get rid of the approximation and see what performance hit it entails? (Probably very little. It'd take longer to get a value from L3...