Game of Life blocks spawned into the map cause the simulation to speed up

Found a bug in R'n'D? Report it here!

Moderators: Zomis, Flumminator

Post Reply
ncrecc
Posts: 107
Joined: Thu Jul 12, 2018 12:59 am

Game of Life blocks spawned into the map cause the simulation to speed up

Post by ncrecc » Wed Mar 27, 2019 3:25 am

Image
I'm guessing this is because the spawned-in blocks don't bother to check whether they're updating "in sync" with the pre-existing blocks, and when the spawned-in blocks merge with the pre-existing blocks, it causes them to update at the same intervals at which the individual clusters would normally update.
(The GoL simulation also grows way further than I would expect here, but I'm guessing part of that is the weirdness caused by the magic ball being considered a live cell without it disappearing.)

filbo
Posts: 357
Joined: Fri Jun 20, 2014 10:06 am

Re: Game of Life blocks spawned into the map cause the simulation to speed up

Post by filbo » Wed Mar 27, 2019 7:56 am

Yeah, looks like spawned-in blocks are on a different time cycle than the original ones. When either set reaches its time to calculate a new generation, the GOL algorithm is run in the neighborhoods of the GOL blocks whose time it is; any newly created GOL blocks will have the same cycle time as the parents which were being checked that time. I didn't look at the code so I don't know if the algorithm scans the entire game board when any GOL block 'ripens', or only those blocks whose timer fired. The latter would make it all run double-speed (then triple, next time the magic ball dropped some in, etc.). If it only scans the neighborhoods of blocks whose timer fired, you'd have weird localized syncopated doubletime...

Looking at the behavior of the (remains of the) glider, I think it's weird localized doubletime. The leading corner of the glider continues to cycle at normal pace, at least until it is eventually engulfed by waves of weirdness emanating from the magic ball.

Holger, how hard would it be to colorize each GOL block according to which cycle time 'pool' it belongs to? (Not necessarily as an actual change to the game, but as a visualization aid here...)

filbo
Posts: 357
Joined: Fri Jun 20, 2014 10:06 am

Re: Game of Life blocks spawned into the map cause the simulation to speed up

Post by filbo » Wed Mar 27, 2019 7:58 am

CE exercise: make a mover which behaves sort of like a mole (probably a slow-moving mole! and maybe a bit more random, not straight-on-until-blocked), laying down a track of GOL tiles as it moves. Each tile would have its own new cycle time, so the overall effect would be spectacularly weird. I think.

filbo
Posts: 357
Joined: Fri Jun 20, 2014 10:06 am

Re: Game of Life blocks spawned into the map cause the simulation to speed up

Post by filbo » Wed Mar 27, 2019 8:00 am

Um. I guess an Android tuned to GOL blocks would be pretty close to that...

filbo
Posts: 357
Joined: Fri Jun 20, 2014 10:06 am

Re: Game of Life blocks spawned into the map cause the simulation to speed up

Post by filbo » Wed Mar 27, 2019 8:14 am

... the extra cycles give it cancer. Just introducing a few blocks at a different cycle causes weird cancerous crud to spread and eventually take over the whole screen.

Post Reply