You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I've seen some long time ago the -next version branch appear, but it seems to not be progressing.
I needed a bit of this, tried to checkout this branch and finish the code but I found many missing implementations and I wasn't sure if everything works as expected.
But I needed such api like in the -next branch so I decided to completely rewrite the whole gl-matrix current stable version into typescript.
I think it might come in handy for the -next branch progress, so I'm sharing it here: https://github.com/adrian-afl/afl-math
How this is done is that I started from the main version JS code and by hand added types everwhere, and then wrapped things in object like on -next branch.
There are some serious changes though - I decided to use x y z object attributes instead of extending Float32Array, this is around the same performance cost, but need to not that raw arrays are 2x faster than this, but the api is not so friendly then. Also this is then Float64 precision because of the "number" type used.
I'm not sure if everything works because I completely skipped the test suite 😞
The package is also ready for NPM package releases, I tested this with Vite based frontends and with backend on Node.js.
I thought this might be helpful for you! Hope it helps with the next big release.
The text was updated successfully, but these errors were encountered:
Hello!
I've seen some long time ago the -next version branch appear, but it seems to not be progressing.
I needed a bit of this, tried to checkout this branch and finish the code but I found many missing implementations and I wasn't sure if everything works as expected.
But I needed such api like in the -next branch so I decided to completely rewrite the whole gl-matrix current stable version into typescript.
I think it might come in handy for the -next branch progress, so I'm sharing it here: https://github.com/adrian-afl/afl-math
How this is done is that I started from the main version JS code and by hand added types everwhere, and then wrapped things in object like on -next branch.
There are some serious changes though - I decided to use x y z object attributes instead of extending Float32Array, this is around the same performance cost, but need to not that raw arrays are 2x faster than this, but the api is not so friendly then. Also this is then Float64 precision because of the "number" type used.
I'm not sure if everything works because I completely skipped the test suite 😞
The package is also ready for NPM package releases, I tested this with Vite based frontends and with backend on Node.js.
I thought this might be helpful for you! Hope it helps with the next big release.
The text was updated successfully, but these errors were encountered: