Bernat

I uninstalled my natively installed version and it removed the udev rules. Now that I've restored them it's working. I'll add a note in the Readme. I'll create a new...

Yes, thanks! I guess this change will be incompatible with previous versions. Would there be a way to keep the compatibility? I'm having an issue with the Gtk3 backend on...

Are you sure that the issue isn't that the forces are reversed? Your change casts an unsigned 32bit value to a signed 32bit value, and that wouldn't have any effect...

These operations can be done with 32bit signed integers, the values they use are 16bit. Indeed, the result will be stored in a 32bit signed variable in 32bits systems. I've...

I don't understand the question. If this PR works, I guess the issue will be multiplying unsigned with signed values. You can write a small test program doing the multiplication...

> I'm saying that previously, you'd fixed 32-bit issues by casting both of those values to (int). Then you changed it in the commit I referenced to only cast the...

This isn't working well for me on a 64bits system. I'll take a closer look and see if there's really a need for 64bits types for this calculation.

I've reviewed the code and a 64bit cast is definitely needed. One question though about the proposed solution in https://github.com/berarma/new-lg4ff/issues/96#issuecomment-2155839482, why use div_s64 instead of relying on the compiler to...

I think I've finally nailed it down. It was a case of me [badly fixing 32bits support](https://github.com/berarma/new-lg4ff/issues/96#issuecomment-2155839482), then [breaking it again trying to fix it properly](https://github.com/berarma/new-lg4ff/commit/2b865afe016b1a25a7a0ff921f94f81f3d1c2969). Your solution of using...

I'll need to do more tests. Apparently, this isn't enough to make it work with 32bits operations. I'll be away from the wheel for most of the summer so it...