3 ms·
This would’ve been a handy article to have read before/during Advent of Code 2018. There were multiple path finding problems I struggled through.
by wycy 8y ago
This would’ve been a handy article to have read before/during Advent of Code 2018. There were multiple path finding problems I struggled through.
- teacpde 8y agoOh man, day 15 is almost a party pooper to me, not because of the algorithm, but the edge cases. Took me ages to debug. It gives me perspectives to game development.
- masklinn 8y agoI don't know that it's quite similar though: advent's day 15 is about replicating exactly the algorithm used by the system designer, except using a somewhat hand-wavy description. Because of the testing methodology/system, it adds a lot of complexity to the task of aggro/pathfinding.
- teacpde 8y agoYeah, you are right, it just happens to remind me day 15. The description is surely not the best, but I don’t think it’s hand wavy, it went great length to explain what happens which I appreciate. What frustrates me is it does not help me to discover certain edge cases. But it reflects reality, there are bugs you can never imagine until you run into.
- wycy 8y agoYeah, my day 15 part 2 is still somewhat bugged. At termination, I print out 6 possible answers, each of which represents an off-by-one error. My correct answer is fortunately one of these: its the round * (elves HP + 3), which indicates that my elves sustain one too many attacks before the game ends. I still have no idea where this extra attack is coming in, but I’ve basically given up on perfecting my day 15 code. It was incredibly frustrating.
- teacpde 8y agoThat sounds very similar to what I experienced, my bug was that a player A plays twice in a round because it steps into a spot yielded by player B which died in the same round. My mistake was bookkeeping in an incorrect (which is only so for a few cases) way. Hope that helps.