Any discussion about Maze algorithms cannot be complete without a reference to the 1982 endless Maze algorithm used in the "Entombed" Atari game.

Many great articles about this can be found like:

https://www.gamesthatwerent.com/2024/01/the-endless-maze-alg...

https://ieee-cog.org/2021/assets/papers/paper_215.pdf

This page has similar visualizations and explanations of multiple (wildly different, in some cases) maze generation algorithms. It’s hypnotic to watch them run.

https://professor-l.github.io/mazes/

This is what I come to HN for. Bravo

Wilson's algorithm is based on so-called loop-erased random walks [1].

[1] https://en.wikipedia.org/wiki/Loop-erased_random_walk

The description of the algorithm is frustratingly confusing.

"Now pick another random black dot to start from and color it white too. From this black dot" from which black dot, the white one?

"single step in a random direction, coloring the new dot white and drawing a line between the two dots". How big of a step is we need to draw a line? Ok, so where not talking about pixels, and where drawing black and white dots on a background of... let's imagine grey?

"backtrack along your path until you’re back at the dot that you were trying to color white" does this algorithm ever terminate in any tractible time?