Rendered at 16:16:56 GMT+0000 (Coordinated Universal Time) with Cloudflare Workers.
Jaxan 9 hours ago [-]
This is very similar to “Mosaic” on Simon Tatham’s puzzle webpage. There it is noted that: “This game is variously known in other locations as: ArtMosaico, Count and Darken, Cuenta Y Sombrea, Fill-a-Pix, Fill-In, Komsu Karala, Magipic, Majipiku, Mosaico, Mosaik, Mozaiek, Nampre Puzzle, Nurie-Puzzle, Oekaki-Pix, Voisimage.”
Single click places an x instead of a sunflower. I solved the puzzle but placed those xes instead of sunflowers and was confused why it doesn't say I solved it. I would suggest making a single click be the main action, ie placing a sunflower.
rezonant 14 hours ago [-]
But you have 3 mistakes you can make, and you spend one when you accidentally place a sunflower. I played through all the daily puzzles and have started into the master puzzles and I'm glad flower is double click, because a misplaced X is free, but a misplaced flower is not.
It would be far too easy to accidentally tap the screen and burn a Mistake, and if its your last one that would be frustrating.
jader201 11 hours ago [-]
Rather than counting mistakes, just allow mistakes, and just not let the puzzle get solved until it’s complete without mistakes.
This is how most of Simon Tatham’s puzzles work, and that works well.
kelnos 9 hours ago [-]
Sure, but then it's easier, and you can just keep clicking around by trial and error without thinking about it, which is a boring way to play.
matsemann 10 hours ago [-]
> This is how most of Simon Tatham’s puzzles work, and that works well.
They btw also have a feature that swaps main and secondary action when toggled, which could work here.
amelius 6 hours ago [-]
Yeah, I made a mistake figuring out the UI. But then I (correctly) solved the puzzle and it said "1 mistake" but it didn't say what that mistake was. This is confusing.
jeremyjh 5 hours ago [-]
The X are like comments, you set those when you've determined a square cannot have a sunflower on it. If you made those double-click and flowers single-click, it would be easy to make a mistake like not double-clicking fast enough and lose to a mistake that isn't really a mistake.
Aardwolf 10 hours ago [-]
It has the option to play a tutorial that explains the controls
nirali-shekhwa 6 hours ago [-]
I agree!
hyperhello 1 days ago [-]
I don’t understand how this is reverse minesweeper when it’s just minesweeper with flowers instead of bombs.
pompomsheep 1 days ago [-]
Yeah it's maybe not the best way to describe it, but my intention is all clues already visible at the start rather than exploration, and adding in flowers rather than avoiding mines.
The main difference is having all the clues visible at the start changes it to a search and deduction logic style rather than reveal and react
tracerbulletx 21 hours ago [-]
It is the right way to describe it, that guy just doesn't remember how minesweeper worked.
imtringued 5 hours ago [-]
The point of reverse minesweeper is that you're supposed to left click the bombs and right click the empty spaces. If you left click the empty squares and right click the bombs, it is literally just mine sweeper.
omoikane 1 days ago [-]
In that sense, this game is similar to Mosaic[1] where the clues are all initially visible, except in Mosaic the count includes all 9 cells since black squares can be placed on the numbers themselves.
You will in particular like this game if you are into conspiracy theories. If one switches the difficulty to the hardest one, HN readers might have some fun with this game. :-)
I agree with you that it's not quite minesweeper, but I also agree with the parent that it's not reverse either. It's sorta just expedited minesweeper
kevincox 24 hours ago [-]
In minesweeper you start with nothing visible and explore while avoiding the bombs. In this you start with all of the hints visible and place the bombs.
However in practice it is very similar since identifying where the bombs are to not click them is exactly the same logic as identifying where the bombs need to go. The only functional difference ends up being that all of the hints are visible upfront rather than needing to discover them.
Bimos 14 hours ago [-]
You don't place the bombs, but still find the bombs. And you still explore by finding x.
kelnos 9 hours ago [-]
It's reverse in the sense that you've been given the numerical hints, and instead of clearing the mines, you need to place them. That feels pretty "reverse" to me.
amelius 6 hours ago [-]
But that's how minesweeper works (except that you have to click to get the first numerical hints but that's not really what would make it "reversed").
A reverse minesweeper would be: bombs are given, and you have to place the numberical hints.
toxik 3 hours ago [-]
Well it is different, you can have multiple solutions in this variant. Classical minesweeper is not like that. Imagine a grid with just a 1 and the rest is empty. There are then eight different solutions. The same situation can't happen in classical, it is simply going to reveal everything except the predetermined mine.
Because of this ambiguity, the search space is much larger here. Source : I play A LOT of Minesweeper.
ryukoposting 16 hours ago [-]
Because "contrapositive minesweeper" doesn't roll off the tongue.
testaccount28 23 hours ago [-]
it reverses right and left click
imtringued 5 hours ago [-]
It doesn't. I first used left click to place the bomb and it told me I'm wrong, there's a bomb there, idiot. Then I started using left click to cross out where the bomb isn't and then I felt like this is just minesweeper, not even reversed.
LearnYouALisp 27 minutes ago [-]
Please enable RMB--the LLM instructions don't say anything about (x), and my button doubles on single often (and drops drag, but we digress)
ltsSmitty 20 hours ago [-]
There were a few boxes in the game I just played that didn't have enough information; e.g. only one number pointed to them and you couldn't reason it out. I think it's the kind of thing in regular minesweeper where it would've auto-cleared it when I opened a box next to it. I didn't take a screenshot but I hope that makes sense - I suppose maybe once you play the game more you'd learn that these are safe x's, but I found it to lack real information. I looked around for a tip or something on it but didn't quickly find anything
pompomsheep 19 hours ago [-]
Helpful feedback thanks.
But yes there is always a logical next step to be made that can be fully reasoned out. The hint (lightbulb) icon will show you this step when you are stuck.
Disclaimer the hint system is definitely not perfect yet but should at least point you in the right direction.
Please do send a screenshot if you find yourself stuck again it would be useful for me to improve the deisng
fallinditch 18 hours ago [-]
I think i experienced the same thing: there appeared to be no logical next step and it was necessary to make a best guess. I probably just didn't see the next step, but I thought it's ok if you have to occasionally guess.
Anyway, great game! Congrats, very entertaining, I reckon you have a good chance of going viral.
jaffa2 20 hours ago [-]
it is possible in normal mine sweeper to have a postion where you cant determine 100% whether a square is a mine or not. In this cases you just have to guess. Try to identify them early as possible since they can ruin the game with an unlucky guess.
For everyone: Whenever you think you have to guess, use the hint feature.
In the more advanced levels you essentially have to "guess" then check whether the guess (e.g. using "pencil mode") leads to a valid solution or a conflict.
Tepix 20 hours ago [-]
First column, there's a 2 in the second row from below that already has two sunflowers. So the 3rd adjacent field must be empty.
MK2k 21 hours ago [-]
The "4" clue forces the solution imho.
21 hours ago [-]
Steve6 21 hours ago [-]
The "2" in col 1 row 5 already has 2 flowers, meaning the empty space in col 2 row 4 must be blank. So a flower belongs in the square above that one and the corner will be blank. No guessing needed.
alex-moon 9 hours ago [-]
Hello Daniel! Do you have a landing page/portal for all these games so I can go through them one by one of a morning (a bit like the NYT games)?
sickcodebruh 16 hours ago [-]
This is great! Did a few rounds, will come back.
A few months ago someone on HN shared Dragonsweeper and it completely hooked me and my wife to the point that we had to agree to break up with it. https://danielben.itch.io/dragonsweeper
18 hours ago [-]
amanzi 20 hours ago [-]
This is very similar concept to "0h h1" https://0hh1.com/ which I've been playing every day on Android for a couple of years now. I had never considered it to be a reverse minesweeper.
That looks like an interesting game, I'll bookmark it. But no - it appears to be a different concept.
adzm 1 days ago [-]
Actually really fun
19 hours ago [-]
pillmillipedes 17 hours ago [-]
this is great! very nice ui, and quite challenging on the high difficulties.
reminds me of 14mv a bit - and judging by the reactions to it on steam you're going to get so many false bug reports from people not used to the long deduction chains haha
(a suggestion to try the hints near the feedback form might help curb this a bit..)
one qol thing I missed is that clicking a number to fill in the neighbours only works for crosses and not sunflowers (and doesn't work for "0" clues)
levanlevi 6 hours ago [-]
the phrase "reverse" confused me (i see somebody pointed out the same)
how many dau you have?
16 hours ago [-]
0x1ceb00da 23 hours ago [-]
Jumped from medium to insane and didn't notice any difficulty bump. Were the puzzles designed by AI, eg arc agi puzzles are trivial for humans but very difficult for LLMs.
mcmoor 15 hours ago [-]
Strangely I also easily solve insane, then I try extreme. First I can solve with difficulty, second I almost gave up and it do need some whole board logic. Haven't tried third.
dyarosla 20 hours ago [-]
Im curious how youve set up the multiple puzzle games you have to generate two daily puzzles each? It cant be entirely by hand or entirely AI driven?
tgsovlerkhgsel 19 hours ago [-]
For any game that is able to procedurally generate a "random" puzzle, you can simply seed a deterministic PRNG with a seed derived from the date. No AI or hand-generation required.
dyarosla 6 hours ago [-]
This is true hypothetically but doesn’t quite work to produce good puzzles like say in Zanagrams without sufficient heuristics/constraints
Dwedit 13 hours ago [-]
Maybe count perfect solves in a separate average than solves with mistakes?
gkfasdfasdf 23 hours ago [-]
I love it. Reverse minesweeper describes it perfectly.
I've been doing these for a long time, which is why this game seemed familiar to me.
ryukoposting 16 hours ago [-]
Very fun. Shared it with my parents and they're into it too.
greazy 17 hours ago [-]
I'm not a big fan of puzzle games but these are great.
Are the clues hand coded?
geekgeekgk 7 hours ago [-]
This is a really nice take on Minesweeper
ttmac 6 hours ago [-]
I really enjoyed it, lost quite a bit of productive time in my work day though!
1 days ago [-]
da-x 1 days ago [-]
Awesome idea
pompomsheep 1 days ago [-]
Thanks! Not perfect yet but I think it has potential. Would love any feedback
Kuinox 1 days ago [-]
I jumped directly to insane and completed it on the first try. It's too easy.
pompomsheep 1 days ago [-]
I've just tried to bump insane up a bit if you could test it to let me know what you think?
I'm struggling to play test the higher difficulties because it above my own skill level haha
vl 12 hours ago [-]
I trained to play insane, usually a just bit slower than average, after using hints for couple times.
I suggest larger levels (yes, scrolling and zooming might be required). After all, minesweeper is the most interesting on the largest level.
kybishop 1 days ago [-]
Insane is still far too easy. For the highest skill level it should be -really- hard. Was still able to solve one in under 3 minutes.
I -love- this idea though!!
pompomsheep 1 days ago [-]
And if you find a puzzle you like the difficulty of could you tell me the puzzle id at the top left please
convolvatron 1 days ago [-]
I guess I am mildly discomfited by the red flower. if the program is going to tell me every time I fail to think through things properly, then my role here is kinda limited to driving the mouse.
pompomsheep 1 days ago [-]
Helpful feedback thanks. I am trying to make it a bit more forgiving for first time players to learn the logic.
Then in easy mode and above you are given max 3 mistakes similar to sudoku.com and other popular logic puzzles
npodbielski 4 hours ago [-]
very addictive :) thanks!
onaclov2000 16 hours ago [-]
Neat game! I enjoyed it!
lioeters 24 hours ago [-]
It's cute and friendly, how it explains the rules. Maybe instead of a red box for making a mistake, the sunflower could "shake its head" to mean "no". Anyway the game feels well-designed with a focused scope, nice work!
mock-possum 9 hours ago [-]
Cool idea but the page doesn’t scroll and the bottom of the puzzle is cut off on smaller res, so - broken for mobile
pompomsheep 9 hours ago [-]
Could you let me know which device please and i will fix
Kuraj 19 hours ago [-]
MSN MESSENGER HAD THIS!
I used to play this with a friend. You would take turns, that is, the mines would reveal themselves in real time, and your turn ended if you missed a mine.
Then the game would call you a slut when it ended, because my friend's MSN was in Swedish and that is the word for "end"
bonoboTP 16 hours ago [-]
No, that was regular minesweeper. The difference in this new version is that all the clues are revealed in the beginning.
Kuraj 15 hours ago [-]
Ah yes - my bad. I fixated on the part where you have to uncover the mines instead of avoiding them.
ymolodtsov 23 hours ago [-]
Honestly looks far more obvious than the original to me.
I like how ai agents make it possible to experiment with different game concepts.
I merged Go with Tetris while brainstorming some ideas.
https://www.chiark.greenend.org.uk/~sgtatham/puzzles/js/mosa...
It would be far too easy to accidentally tap the screen and burn a Mistake, and if its your last one that would be frustrating.
This is how most of Simon Tatham’s puzzles work, and that works well.
They btw also have a feature that swaps main and secondary action when toggled, which could work here.
The main difference is having all the clues visible at the start changes it to a search and deduction logic style rather than reveal and react
[1] https://www.chiark.greenend.org.uk/~sgtatham/puzzles/js/mosa...
> https://store.steampowered.com/publisher/divide-the-plunder
I particularly like "Mosaic of the Strange"
> https://store.steampowered.com/app/3444650/Mosaic_of_the_Str...
You will in particular like this game if you are into conspiracy theories. If one switches the difficulty to the hardest one, HN readers might have some fun with this game. :-)
try this one in html5: https://hojamaka.com/game/mamono_sweeper_h2/html5/en.html
However in practice it is very similar since identifying where the bombs are to not click them is exactly the same logic as identifying where the bombs need to go. The only functional difference ends up being that all of the hints are visible upfront rather than needing to discover them.
A reverse minesweeper would be: bombs are given, and you have to place the numberical hints.
Because of this ambiguity, the search space is much larger here. Source : I play A LOT of Minesweeper.
But yes there is always a logical next step to be made that can be fully reasoned out. The hint (lightbulb) icon will show you this step when you are stuck.
Disclaimer the hint system is definitely not perfect yet but should at least point you in the right direction.
Please do send a screenshot if you find yourself stuck again it would be useful for me to improve the deisng
https://imgur.com/a/idfj1my
edit: I'm wrong. Thanks steve.
In the more advanced levels you essentially have to "guess" then check whether the guess (e.g. using "pencil mode") leads to a valid solution or a conflict.
A few months ago someone on HN shared Dragonsweeper and it completely hooked me and my wife to the point that we had to agree to break up with it. https://danielben.itch.io/dragonsweeper
Online example for the curious: https://raetselwelt.online/en/puzzles/binoxxo/
reminds me of 14mv a bit - and judging by the reactions to it on steam you're going to get so many false bug reports from people not used to the long deduction chains haha
(a suggestion to try the hints near the feedback form might help curb this a bit..)
one qol thing I missed is that clicking a number to fill in the neighbours only works for crosses and not sunflowers (and doesn't work for "0" clues)
how many dau you have?
I've been doing these for a long time, which is why this game seemed familiar to me.
Are the clues hand coded?
I'm struggling to play test the higher difficulties because it above my own skill level haha
I suggest larger levels (yes, scrolling and zooming might be required). After all, minesweeper is the most interesting on the largest level.
I -love- this idea though!!
Then in easy mode and above you are given max 3 mistakes similar to sudoku.com and other popular logic puzzles
I used to play this with a friend. You would take turns, that is, the mines would reveal themselves in real time, and your turn ended if you missed a mine.
Then the game would call you a slut when it ended, because my friend's MSN was in Swedish and that is the word for "end"
I like how ai agents make it possible to experiment with different game concepts.
I merged Go with Tetris while brainstorming some ideas.
https://goban.molodtsov.me/