This being my first post, there’s a big yellow banner on the side suggesting comments should be constructive, and lucky for you because otherwise I’d go off on this puzzle. I joined just 2 days ago and I’m only 3 steps into the “algorithms” track, having just finished a few puzzles with “loops” (not making this up, that’s literally how they’re tagged) and 2 “medium” difficulty puzzles, one which I solved using a smidge of recursion plus eval (which wasn’t disabled so I guess it’s legit) and a supposed trig problem for which I didn’t need any trigonometric functions nor even square root from the math library. So now I’m on the next milestone which is puzzle of the week, and correct me if I’m wrong but there’s only one puzzle of the week I can be working on right now and this is it this week. Well working on it is what I’ve been doing, looking at those pretty pictures with dots and lines and painstakingly coding up a method of counting the little guys all the way around, only to find that after a bit of effort to correctly exclude the boundary and such, it works great on the examples given but times out on the other two tests. So no problem, it’s simply a slightly more complex optimization problem, and anyway I’m pretty handy with a keyboard or at least the dexterous half, so how hard could it be to get one more of countlessly many achievements, on just step 3 of the algorithm track? I optimize the sh*t out of it (sorry for swearing but I really did do a number on it), using set operations on column masks instead of looping thru each Moore neighbor, and yes getting the right answer on my machine at home but finding that it still takes a few minutes to run, only to finally come here and see that Froot Loops are for kids and the way this one has to be done, silly rabbit, is by knowing some unnamed theorem (and I mean unnamed here in the forums, while in the problem statement not so much as mentioned or hinted) making explicit use of the fact that all the vertices are integers as probably its strangest and most coincidental condition. To make the situation more of a downer in my case, I actually knew about this theorem from high school geometry, but to my memory it only worked on convex polygons, not on any simple polygon like the whack job going on in example deux. So I guess all of this is to say I’ve got a bit of a sour taste in my mouth from this experience in contrast to the otherwise very polished content on this site, and I hope that’s what was intended for me to learn my lesson about trying to figure stuff out on my own instead of asking Jon Skeet. But if not then maybe, just maybe suggest that actually counting the dots might not be the best approach, if only because this is the singlest problem that can get past that single hurdle at the moment. Thanks for listening, and regardless of how simple it turns out this is to implement (or probably worse because of how simple), I’m far too burned to care about this puzzle anymore. Maybe I’ll come back and try the one next week, we’ll see. Edit: Whatever, I’m over it, after completing in 10 lines of code with 100 lines commented out,
and giving it the one full star it deserves.