Indie game storeFree gamesFun gamesHorror games
Game developmentAssetsComics
SalesBundles
Jobs
Tags
(6 edits)

great tool!

I think I found a bug. Load v0s0r1w30h3l30a8b11c9d9pZw9Z4Z4Z4AZ3AZ3A and take one step. The expected result is a 2x3x2 block, however it is missing two cells in a way that breaks the symmetry of the original figure.

(8,11,9,9 is conway, adapted for simulating warping 2d in warping 3d. This particular example seems to work fine for heights of 1, 2 and 4, the bug occurs with height 3.)

edit: Different example:

glider that stops when warping:

v0s0r1w30h1l30a8b11c7d11pZjyZ5Z2AA

functional counterpart with slightly different length:

v0s0r1w30h1l31a8b11c7d11pZjxZ5Z2AA

(1 edit) (+2)

Edit: I have released a hotfix which I believe might have fixed the issue!

Thank you! I looked into this and I think the reason for these bugs are due to your graphics card rounding some floating point operations differently. If you know, what graphics card are you using?

I haven't been able to reproduce the first and third scenarios that you sent but I was able to reproduce the second scenario where the glider doesn't wrap correctly.

Have you been able to reproduce these bugs in any scenario where the size is an even number in all directions?

This problem might be impossible to fix if some graphics card simply don't give accurate results but I will see what I can do.

(+1)

Yup, seems like you fixed it, all the scenarios in my comment work correctly now. Good job!

For the case that needs clarification, the third scenario was never buggy for me.