Just finished 80: Wrong game. I thought it was great. More thematic levels early on might be really good for keeping newcomers playing

@NathanMerrill Thanks, that used to be my favorite level, but a couple of people pointed out that at least 2 steps in that level are... too hard, I guess.

There are some hard points

there was one step in deduction where I had to add up several rows and compare them with tiles

I think opening up the lower right early on, and then much later dealing with the lower left both require you to figure out that a certain tile can't be a mine (or must be a mine? I forget)

but that's my favorite logic

That's a different way to do it, that I didn't do; you can also try out the individual tiles, and there's one key tile in each step that will give you all the info you need

yeah, I know exactly which points

I think the 100 levels are pretty much fixed now, but I can add some more fun levels in the bonus set on the side

My only real goal going forward with the game is to fix bugs, shore up the code, and add somewhere between 40 and 100 more levels over the next year. Then I will probably stop development; anything further would be for a sequel

@EricTressler Are you planning a sequel? What do you envision would be significantly different?

@DJMcMayhem Not right now; I think I might do some research in my spare time after this, or some other project. This has been very time-consuming.

It's hard to imagine what you could add to it to make it worth being a different standalone

well the hexcells games are all standalone and they could as well have been one game

It would be interesting to try other neighborhoods, but I can do that in this game if I decide it's worth the burden it places on the player (probably not).

@MartinEnder True, but Hexcells is really popular, so maybe a smaller game is better?

@MartinEnder Have you ever seen

yeah, actually I have

My dream is a metroidvania with regions populated by enemies that are trained in a similar way, with more training as you go on. And some kind of SHODAN AI pasted on for story

But every time I play with neural networks, I end up trying to solve actual problems instead of working on something like that

heh, that could be quite fun. although, as much as I find procedural generation really interesting, so far I have yet to see a game where procedurally generated content is as captivating as handcrafted content. I feel like enemies which get trained instead of having fixed patterns might somehow feel like they have less character.

Maybe; there are like 5000 steps missing in my plan

@EricTressler 2:01 gave me this hilarious mental image of giant T-Rex, furiously hopping towards it's prey lol