https://adventofcode.com/2021/day/11
Another grid problem like Day 9. I chose to implement a helper function to calculate neighbors without padding the grid like I did in day 9. That might have been faster to code, however, this worked very smoothly.
Shockingly the real grid was the same size as the sample grid.
In the end the first part was similar to part 2 of day 9. A simple recursive DFS with tracking to implement "flashing" chain reactions and finally a second pass through the grid to reset flashers to 0 and count the total number of flashes.
The second part was just running the interation past 100 days to see when all jellyfish hit 0 at once. I had an off by one error on this one, that probably cost me a 100 places.
Reader, we animated the grid:
https://scratch.mit.edu/projects/615751043/
This combine visualization code from @whaleandbeaver
and a fair amount of simulation code. I uses two different encodings for points an index 1-100 and [x,y] in 0..9 x 0..9 and conversions between the two led to so many painful debugging sessions. In retrospect, I should have returned neighbors as a list, and then kept a queue, looping through every point that needed to be visited in the Flash update rather than trying to use recursion since I had to keep two stacks to make that work in the end.
there were also definitely some hilarious bloopers when the visualization and the computation got into a bad state
