Giving away:
- 4x Piczle Cross Adventure(no Turkey, Switzerland and Argentina)
- 2x Suicide Guy Deluxe Plus
- 2x Smoke and Sacrifice(no China)
- Diluvian Ultra
- De'Vine: Heavenly Acres
- De'Vine: The Card Battles
- De'Vine: World of Shadows
- Once upon a Dungeon II
Comment has been collapsed.
Thank you for the clever event :)
Once upon a Dungeon II link needs a final 0, 2204730.
Comment has been collapsed.
Blacklist check
99% of my blacklist is people who I think are using scripts to autojoin raffles.
If you think you don't deserve to be in there, feel free to let me know.
Comment has been collapsed.
what the heck
how did you know!!!!!!!!!!!!!!!!!!!!!
Comment has been collapsed.
Oh, so the message is both the seed and the reward? Just learned something, cool!
Thanks!
Comment has been collapsed.
Hm, yes, you are correct, of course.
Another question I have: can the site actually generate badly formed nonograms (like, with multiple solutions)?
Comment has been collapsed.
No, the method to generate a nonogram is not randomly putting dots and then X out the empty ones. It's a known algorithm that ensures that the created nonogram has no multiple solutions (I'm not sure but I'd say it's impossible to make a nonogram with more than 1 solution), and more importantly, you never have to guess. Simply following logical steps would lead you to the solution without doubts. If you're stuck, you're not seeing a logical step; or in my case, you were absentminded and already misplaced a dot or an X.
My only gripe with RosimInc's Nonogram Generator is that you can't give it a "density". The more dots it has, the more easy it is; the less dots it has, the harder it is. Of course there's an upper limit on density where solving it is boringly easily, and a lower limit where the algorithm can't create a "no guesses" nonogram. So with this generator you can find easy, quickly solvable ones, even big, or very hard, slow to solve ones, as I've found in the last months when nono's have been popular again.
Edit: someone as despicable as Gru, but not as carismatic, and probably cruel to children and kittens, has shown me with good proof that indeed there are nonograms with more than one solution. I don't think the algorithm that RosimInc used could produce that kind of nonograms, but well, I can ONCE AGAIN be wrong. I'm used to it, so no hard feelings towards this pergsonk.
Comment has been collapsed.
Simple proof that some nonos have multiple solutions:
Imagine creating an nxn nono by filling in only two spaces - one at 1,1 and the other at n,n. The resulting puzzle would have a second solution where 1,n and n,1 are filled in instead.
Comment has been collapsed.
Yeah, what really prompted me to ask this was the "you test it first" statement in the OP. Here is a challenge: can anybody generate a "badly formed" (meaning, multiple solutions) nonogram with RosimInc's generator? If yes, share it with us! If no example can be found, this would confirm your statement that the algorithm itself eliminates the bad ones.
I see your edit -- I knew multiple solutions to be possible because my favorite Nonogram app (CrossMe, for Android; the pictures there mean something at the end, makes it more charming) very rarely had them -- which they promptly fixed with updates, but still, I have seen them.
Not sure if this is cruel Gru's example, but a 2x2 nonogram with hints "1" in each line and column has two solutions... I imagine "low densities" have this problem more often then "high", as you say.
I actually have this dream of solving a few mathematical questions about nonograms, like:
i) What is the minimum density of a nonogram to guarantee it is well formed? Is it 100%? How does it change with the size of the nonogram?
ii) What is the average density of a well-formed nonogram (using a probabiliy distribution that assigns 50% to each pixel, then discards the bad nonograms)? Does it approach 100% when the size grows, or does it approach some other number?
The first question might have an easy solution, maybe? Anybody knows of high density nonograms which are bad? Now, the second question... whenever I start to think about it, I realize it is way beyond my grasp... :D
Oh, crap, I wrote a wall, didn't I? Sorry! :D
Cheers!
Edit: OH WAIT, as I was writing this I realized the answer to (i) is clearly 100%, with a pretty simple example!!!! I will leave the solution to the readers! :D :D :D
Comment has been collapsed.
Your 2nd question is quite interesting, however I think that most nonograms could easily be tweaked to become unsolvable, by adding that X pattern of White-Black, Black-White (surrounded by white of course)
So it doesn't necessarily imply a big or small density, lots of puzzles can be unsolvable if they feature that pattern.
I have seen a nonogram generator that basically 'solved' that problem by marking one of the cells with its answer already so that there is only one possible solution. I don't think I will go that route for the tool but it is one possible solution.
If you want a high density puzzle that is bad, I can give you this one (image below)
The two 1's make it impossible to have an unique solution, and out of 400 cells, only 6 are white.
And that is a 20x20, but we could go much higher to essentially have a density approaching 100%
Comment has been collapsed.
Yes, that was exactly what I thought for question (i)... But what you say is super clever, that it might help with (ii) as well, right? I mean, for every (n-3)x(n-3) "good" nonogram, you can easily use this trick to make both good and bad nxn nonograms with nearly the same density... This just might imply that the answer to (ii) is actually 50% when n goes to infinity... well, maybe, there are still many details to make this into a formal proof, but it is a neat idea.
Comment has been collapsed.
I am looking at potential ways to parameterize the difficulty of the puzzles, so that for example we could get a small, but difficult puzzle or a big, but easy puzzle. The density is one of the things I'm looking at, I might actually propose a few different options so that people have the control over what kind of puzzle is generated.
I haven't had a lot of time to further develop the tool lately but it's one of my priorities for the next updates!
Thanks for the feedback!! <3
Comment has been collapsed.
The answer to (i) is less than 100%. If you put an 1 in column 1 and another in row 1, you get a density lower than 100 and its well formed. I'd say the answer is high density but it's not. In a grid where width and height are odd, make a checkerboard pattern starting with a dot at the top-left corner. I think, without thinking much, it's well formed. The density would be around 50%.
I find this very interesting and fun, but my maths base is zero. Like, 2+2=5 for large values of 2. That's as much as maths I can do.
Comment has been collapsed.
That's great! I hope you didn't get my point as negative criticism. It was just a comment that somehow determining the difficulty would make the generator amazingly better. I'm glad you're working at it, I hope it doesn't give you much headaches!
And of course, I'll take this chance to thank you for the fun you've created with the tool.
Comment has been collapsed.
Haha don't worry, I want to make the tool as fun to use and solve and I love when people share ideas and criticism :)
It does give me headaches sometimes, but the good kind! Especially that since the code is getting bigger and bigger, I need to think of better ways to organize it and I also want to make sure that older puzzles will still work even if I change the solving/generating algorithms ^^
Comment has been collapsed.
Creator here, the tool generates random nonograms, but there is a check done to make sure the solution is unique and can be found with basic logic (without guess-and-check or trial-and-error). If the generated puzzle doesn't comply, then I generate another, until I get a valid puzzle ^^.
Not the most optimized approach, but it works reasonably fast :)
Comment has been collapsed.
Oh, wow, this is awesome. But wait -- if you have the checking algorithm in place, would it be hard to make the user choose whatever dots they want in the puzzle, and the algorithm would then check if it is good or not? Then people could draw pictures, houses, dogs.... 5-letter codes.... you know, the usual. :D :D :D
By the way, I never said it, so: thank you for the site (and all that goes with it, the Weekly thread and all) -- nonograms are awesome, and you can see how many people in the site love to create and solve them. You rule. :D
Cheers!
Comment has been collapsed.
The checking of puzzles in itself is straightforward, so nothing would prevent the solver from checking custom-made nonograms.
But making the manual nonogram creator, that's the part that's a bit trickier ^^.
For now my code isn't flexible enough to be reused as-is to create custom nonos, so I'm rethinking the approach so that I can reorganize the code and make it as reusable as possible, while allowing for new types of generations or solves.
Also if a puzzle isn't possible, I would like to be able to highlight why it is not, so that the creator can correct little details here and there, that will be a challenge too ^^.
The other thing is that the solver is a little slow at the moment (one of the reasons why puzzles are limited to 25x25), so I would also like to improve the logic.
At first I only created the random generator because I personally have no creativity, artistic skill nor patience to create the nonograms by hand. But I know that many people here do, and it's the next priority for me to allow custom nonos :)
I just don't know whether it will take me weeks or months hahaha :P
It makes me really happy to see the positive response everyone has towards the tool, seeing the lovely comments and even puzzle threads using the creator :)
Thanks for your lovely comment! <3
Comment has been collapsed.
I wonder too... 19 hours later, I see people have already delved super deep into the topic. 😆
Comment has been collapsed.
If it does, I'm glad it can't be done with bigger nonograms
Comment has been collapsed.
I've done several 25x25 ones, and some took me several retries, each one pretty long. So even if it's restricted to 25, I still consider that it might render a quite difficult nono.
Comment has been collapsed.
hmm, that's interesting, and shocking at the same time.
Comment has been collapsed.
Oh my word, this was so smart! :D
Comment has been collapsed.
That's right, I'm tired of making puzzles. How about you make one for a change, and see what it's like!
Well... I can't expect too much from a first timer like you. Let's start with something simple... how about a nonogram?
Yeah! They're popular, and easy to make with this tool: https://rosiminc.github.io/sg-nonograms/creator
I'll even give you a tutorial.
A 5x5 should be good enough to get started. You can use "potato" as the secret message, just to see how it works.
Now your puzzle is ready, time to post it... WRONG, you test it first.
So go on, solve the nonogram, see if it's really doable.
Done? All right, now you're ready to make a real puzzle!
After you get used to it, you can start using other tools, like this one, it looks fun: https://lockee.fr/o/J8GURjYt
Oh, you can ignore that puzzle specifically... No no, really!
Comment has been collapsed.