advent-of-code VS advent-of-code-2023

Am I going mad, or has anyone else noticed that an answer they submitted yesterday that was rejected is now being accepted? Is this the first time that the "My answer wasn't accepted - must be a bug in AoC" meme might actually apply?? I was pulling my hair out yesterday trying to find out why my answer for part 2 wasn't being accepted, and I've tried inputting the same answer today and now it's being accepted - I've also checked against another solution that was upvoted a lot on this thread that gave a different (i.e. presumably incorrect) answer, so I'm wondering if there was actually a bug in whatever AoC uses behind the scenes to generate the correct answer, at least when applied to my data? Anyway, here's my solution - I start at the top left loop node (which must be an F by definition) and then walk the loop clockwise (at least, clockwise initially) from there, doing DFS on each node that is to the right as I walk the loop (i.e. right from the walker's perspective, not right in the grid) and adding nodes that the DFS finds until it reaches a loop node. I've not seen anyone else implement this so thought it was worth sharing