mystise

So it turns out this is because Open Simplex surflets have a slightly larger radius than one simplex, and as such vertices outside the current simplex have to be taken...

@vallentin It's been a while so I'm a bit fuzzy, but I'm pretty sure I abandoned OpenSimplex in favor of SuperSimplex EDIT: or, as per my comment above, just revert:...

As a possible alternative in the meantime until further work is done, Super Simplex (added in https://github.com/brendanzab/noise-rs/commit/051f8f7180bf0daf64d71bf22dc8c2337fa0b70e) is scaled to [-1, 1] as close as is possible with f64 precision....

I can't speak for any of the other generators, but I can at least speak for SuperSimplex. While I don't have formal mathematical proof, I did use Mathematica to find...

This is perhaps due to: https://github.com/Razaekel/noise-rs/issues/141 manifesting weirdly in 4D

I permit the license of my contributions to the BuildCraft project to be changed to the Mozilla Public License, v. 2.0.

I've reproduced this one with interesting results, the bug only appears when the wrapper header is **not** using absolute paths. ``` $ cat wrapper_tmp.h #include $ bindgen --no-recursive-allowlist --allowlist-function SDL_IsTablet...

Are those accessible by the CLI? My current experiments have been purely using the CLI, so I don't know about the API surface.

Yeah, adding a per-macro override (exposing `int_macro` to the CLI) would be a nice feature to have. Also the docs in CLI for `--fit-macro-constant-types` currently says `Try to fit macro...

Information for future people who get just as annoyed at this as me: This *also* occurs when the rename would create a naming conflict (say renaming a variable from `abc`...