@MartinEnder knowing you enjoy it pretty much guarantees its a game I'd enjoy. Bah. I don't really have the time right now.

@EricTressler Wait, a 75 percent discount on a three dollar game?

Interesting

@DJMcMayhem Sorry didn't see your message. If you see me on Steam, I'll be way more responsive on that than I am on chat here.

@EricTressler 56 (Spot) was nice :) ... I think I generally prefer smaller levels that explore somewhat new ways to combine the hints over the larger levels like 55 (Emergency 2)

@MartinEnder The levels tend to be a bit bigger in the second half of the game, unfortunately. I don't think it gets too out of hand, but I guess I'll find out how you feel about it soon enough.

yeah I can imagine that it gets tough to set up the kinds of harder puzzles you'd want for later levels with small grids

Yeah. I don't want to spoil anything, but I can say more after you've beaten another few levels.

@EricTressler just finished 60 (Windmill). I really liked this one. It felt quite tough but very satisfying to figure out. I don't know how linear these are, but I got to a point where I needed to figure out where the three yellows go for three different reasons, so that I could uncover all the other yellows. It took me quite a while to figure that out, but after that the rest was just easy enough to feel very rewarding, so nice work on this one! :)

@MartinEnder I saw someone actually count grey tiles at that exact point, and that apparently also works

I tried that as well, but it didn't work out for me, maybe I was overlooking something.

But I take that to mean the puzzle is sort of designed around this step as a bottleneck?

@MartinEnder The mines aren't laid down by hand; the bottlenecks happen where they happen, but I kept the good ones and tuned things and sorted them by difficulty. That does seem to be a bottleneck, but sometimes there is a tricky way to proceed if you miss the "standard" solution.

ah okay. I honestly have no clue where I'd begin designing puzzles like this.

Yeah, me either. I made some of the early ones, but I couldn't build any of the difficult ones. The upside is that I got to play it too, even if I can't take personal credit for how all of the puzzles turned out.

so you came up with the layouts and colour distributions and then generated mine placements until the hints were solvable?

no, I came up with the layouts and colors and placed the column hints, and then the algorithm adds and shifts mines (and adds "?"s), each time leaving it solvable and scoring it for difficulty

it's a hill-climbing algorithm, more or less, with some difficulty metric as the fitness function

ah, interesting

The difficulty is extremely sensitive to small changes, which is why I think I couldn't design them by hand. You get these chains of implications that would just fall apart if anything moved.