Brad Allred

wont this break Python 2.x support? I think you need a future import for `print` If the intent is to nix Python 2.x then you should also replace the `simplejson`...

I feel like we've touched on this elsewhere at some point (I think I wanted to drop `visibility=hidden` from `core`), but the issue has nothing to do with `Python`. There...

I'm not getting side tracked, I'm sharing my experience from decades of cross platform development regarding symbol visibility issues. I'm suggesting that we identify the problematic symbols and ensure they...

> here we currently have a clear smoking gun which is? > And we worked fine on FreeBSD for ages, so it's not at all clear if that's an issue...

I'm having difficulty conceiving `NormalizeDeltas` producing a meaningful difference between `float` and `double`. If true tho, you should be able to repro on another system by changing it to `float`...

It might help if you elaborated a bit. Have you ruled out `-ffast-math` and `-frounding-math`? It would still be nice to know the specifics so we can consider implications beyond...

these type of comparisons seems suspicious `dx == 0.0`. I believe thats the sort of thing thats not going to be well defined with `fast-math` and also that `0.0` is...

it looks like there is a bug here: `dy = STEP_RADIUS;`. If I'm not mistaken, that should be `dy = STEP_RADIUS * 0.75;`, no?

> `std::hypot` generates some diff noise when using it, in the less significant parts of course but I better don't try to analyze the numerical impact here now. If the...

> Sounds somewhat counter-intuitive to fix a problem of small errors with even bigger errors? I don't know why you're insisting the problem is one of precision. just because the...