I would also test this zoomed out and with the player running, if you chose to add this later your going to have some issues. Why not preGen more chunks ahead and cull the rendering you should get better results
I think that makes sense as a stress test, and I'll probably do something like that at some point, but I'm trying not to get too bogged down on things at this stage. If I do end up needing faster generation at some point I have options with this approach, even potentially to the extreme of using WASM to run the WFC calculations in Rust or something if I really want. But I'm also planning on polishing/simplifying the tile sets I am using as well, which should also help speed up the generation.
Why generate the entire chunk tiles all at once? It seems like you only really need to compute one vertical slice of the tiles at a time. Is it a concern that backtracking will not create consistent terrain?
Now that I've added the ability to use "seed" tiles for creating smooth WFC transitions, an approach where I generate just one column (or several) of tiles would also be possible, though I'm not sure if it would offer any benefits over what I'm doing now (except perhaps that I potentially might be able to avoid web workers, but at least at the moment I don't have any concerns with that approach). As for backtracking, I will eventually use seeds so the RNG works out the same for the generations each time.
definitely very interesting and engaging keep it up man
This was really concise and kept me engaged the whole way. Great job. I’d love to see more!
Keep up the good work, I wanna start game dev too and these videos help to learn many ways to make a game
this is awesome dude ima pick it up when it comes out
I would also test this zoomed out and with the player running, if you chose to add this later your going to have some issues. Why not preGen more chunks ahead and cull the rendering you should get better results
I think that makes sense as a stress test, and I'll probably do something like that at some point, but I'm trying not to get too bogged down on things at this stage. If I do end up needing faster generation at some point I have options with this approach, even potentially to the extreme of using WASM to run the WFC calculations in Rust or something if I really want. But I'm also planning on polishing/simplifying the tile sets I am using as well, which should also help speed up the generation.
Underrated channel
Why generate the entire chunk tiles all at once? It seems like you only really need to compute one vertical slice of the tiles at a time. Is it a concern that backtracking will not create consistent terrain?
Now that I've added the ability to use "seed" tiles for creating smooth WFC transitions, an approach where I generate just one column (or several) of tiles would also be possible, though I'm not sure if it would offer any benefits over what I'm doing now (except perhaps that I potentially might be able to avoid web workers, but at least at the moment I don't have any concerns with that approach). As for backtracking, I will eventually use seeds so the RNG works out the same for the generations each time.
@@typedpixels Sounds reasonable. Looking forward to it.
Do your parents know that you are gay?
Do yours?
@@someyetiwithinternetaccess1253 Mine ne knowe him.