[_Microft]

Thanks, cool game! Three lives is pretty challenging on larger fields with lots of unavoidable guess by the way.

Since there's no email adress to be found nor PMs open on Twitter, here is a bug report: there is a mine next to a field of revealed, blank hex fields which shouldn't be possible if blank fields mean that zero mines border to them.

The game is still open, let me know if I should look up a seed or extract whatever useful information for you.

Screenshot: https://i.imgur.com/H5YhfqA.png

[floatrock]

> Three lives is pretty challenging on larger fields with lots of unavoidable guess by the way

Agreed. On boards like Flower where there's a long thin strip of tiles, it's very easy to get to a point where it's pretty much a 50/50 guess, no way to logic out of it.

I like the idea of the unique board shapes, but if there's no practical way to win because you just keep hitting wall after wall of 50/50, well, that kinda takes the fun out of it.

Maybe need to think beyond classic minesweeper mechanics for a way around this to keep it interesting?

[_fp5w]

Hey, thanks for the bug report and taking the time to SS this and track me down here. I'll trace it in memory logs to be sure, but, I think it's on a fence between a bug and a not-bug. Time to torque all the bolt heads for sure.

The recursive clearing routine honors your bad guesses and does not attempt to correct you. Confusing part is that the interface doesn't explain this when it proc's and a detonation occurs. But the maybe-new bug part here you found is likely that an edge detonation shouldn't trigger a clearing routine but I think it did in this case for some reason.

Part of this sort of thing tends to normally happen if you either accidentally do a near-simultaneous right-click and left-click on some sensitive mice, or TCP network latency causes commands to get transmitted in a fast burst. In that latter case, I don't think order of the clicking messages is messed up by network conditions, but, ... I'll double check that to be sure and keep an eye on it, so again, thanks for taking the time to report this.

My alpha tester has finished most of the large maps without hitting a single mine, after many grueling attempts. She started with the flower one and blew out her boyfriend's eardrum screaming about it when she won. We heard the scream bleed through on discord and then there was a man down in Overwatch.

Right now all the shaped maps are manually rasterized. I'm halfway to switching that process to a dynamic on-the-fly one since to my surprise and delight it turns out it's not computationally intensive. But all other than beginner/inter/expert are lazily set at that expert density, vaguely derived from the classic minesweeper's field percentages (20.625%).

The next update will let you select more granular game options, like exact difficulty per map, map tile size (dynamic rescaler), etc. Options right now are to play with multiple players, so you then have 3x lives per player, or of course cheat and invite yourself to the game in a different tab and finish the map ;)