That was fun until I got to the point where no progress could be made and I had to undo a whole bunch of times to get to a workable configuration. Perhaps add a notification of some kind that I've gotten myself in that situation, rather than letting me kill a bunch of time solving an unsolvable puzzle. Still, very enjoyable!
Same here. On yesterday's puzzle I got into an unwinnable state on my very first move, using the topmost word (spoiler: I made "hut" instead of "out").
lmao same. actually a really cool fun/concept it's definitely wordle popularity caliber, but once i got to the last 3 words and ended up in this scenario and the hint button said that i was like -_- owned.
not sure what the right game experience would be for that. a notif that says "You can still solve more words but you'll never solve them all!" doesn't quite work here, because it's sort of saying "there's only one _right_ way to win, but good luck figuring out the right order". Still, it would be better than me finding that out at the very end.
it would probably be pretty important to design levels so that the unwinnable states can't happen early in the game, but it's getting a little abstract to think about at this point. sort of brings me back to that unblock it game from the old ipod touch days.
maybe a percentage chance of solving puzzle tracker that updates a bit randomly slow so you don't necessarily know right away that you made a mistake, although it would have to be a bit weird, for example when you start you are not at 100% of solving puzzle.
This was really fun. I actually went back to the beginning of the archive to find more puzzles to solve. I think it would be nice if when you have solved one of the historical puzzles it would give you a button to click onto the next day's puzzle.
I actually never came across the situation where my grid was unsolvable. As I was playing I was wondering what kind of algorithm you used to come up with grids that have no possible dead ends. Since this is clearly an issue, I wonder if it's possible to come up with puzzles that always have a solution similar to how the algorithm for creating Sudoku puzzles always guarantees it can be solved without guessing.
It’s stuck at “Loading…” on both Brave with no extensions and Safari with adblockers. It was only after I tried it on Firefox without anything that I realised the cookie banner is the issue.
In Firefox with ublock origin, I got in the console:
"Loading failed for the <script> with source “https://uncrossy.com/js/cookieManager.js”." looks like a false positive in the ad blocker. Allow-listing the site fixed it.
Fun! On first thought, I'd prefer knowing when I'm in an unwinnable state instead of having to keep clicking the hint button.
Also, the site worked for me in Chrome but doesn't work in Firefox (145.0.2). Do `window.cookieManager = ...` (or even `var cookieManager = ...`) instead of `const cookieManager = ...`. This goes for all variables in the global lexical scope you intend to share across source files.
The website says "I basically need to design puzzles in reverse and have built a set of tools to help me with that." Does that mean that each day's puzzle is essentially built by hand? It seems like an interesting and non-trivial search problem to automatically generate puzzles, given a dictionary.
Uncrossy creator here. Yes, I design each by hand using a highly custom editor that I built to help with many of the common challenges. People keep asking me if it's a job that AI could do, but I think it would be really hard to get right. There's a lot of craft in building the shape of the puzzle and picking words that allow as many spaces as possible to get filled in.
That's super cool! Thanks for explaining. Just to clarify myself - when I was thinking of automation I was not thinking of AI, or at least not how its popularly understood these days. I was thinking more like how 2D mazes can be automatically made by non-AI methods. Uncrossy would need a kind of constraint solver that can navigate words lexically and the spatial intersection of words (and their possible slides). I'm certainly not telling you anything you don't know. There are some algorithms for these (general) kinds of constraint satisfaction problems and they long pre-date "AI", but I think the application to uncrossy and refinement would be quite challenging.
Very nice. Easy to accidentally cheat, however. Shift a word to an invalid position, but right click instead of letting the mouse up event fire. Then shift the word back to the original position: win!
This is really, really cool. I think telling me how many moves I have to go back in the hint was absolutely a must-do, and shouldn't cost two hints... Second-guessing every single move I made would be insane, but knowing I had to go back seven, and pick something different than the last thing I restored, that worked fine.
It's easy to assume making a word disappear is always the right choice, but you forget it changes the word it leaves behind as well. Very clever.
It does have the same quirk Wordle had that bugged me: Treating browser storage as useful in our multidevice world.
> It's easy to assume making a word disappear is always the right choice, but you forget it changes the word it leaves behind as well. Very clever.
It's also easy to make the opposite assumption, that the goal is to change the other word. I initially felt weird about changing from a letter at position 3 to the same letter at position 1, but eventually realised that the goal is just to slide the word around, not necessarily to make a new word.
Fantastic work, very fun !
I actually only ran into the dead end scenario right until the last few words so not a frustrating first experience. But reading other comments maybe a setting to prevent the player to take a route that ends up unsolvable would be great. Kinda like the "Normal" and "Expert" Modes in worldle
Seems like the same idea as treating your LLM as a compiler.
When you write code in Rust and user your compiler to produce an x86 binary, you don't maintain the binary. If you want to make a change, you toss away the old binary, change your Rust code, and recompile.
That concept has been floating around a few places recently. It is wishful thinking. Your userbase won't accept constant small changes in UX, functionality, or the same bugs returning every time you update the production code. And that is what you will get if you run a non-deterministic compiler that hallucinates.
Was a bit disappointed when I almost "solved" it but couldn't solve the last 2 words, finally clicked the hint and it told me to undo 12 times.. would have preferred if there was a warning earlier.
That was fun until I got to the point where no progress could be made and I had to undo a whole bunch of times to get to a workable configuration. Perhaps add a notification of some kind that I've gotten myself in that situation, rather than letting me kill a bunch of time solving an unsolvable puzzle. Still, very enjoyable!
There is a "Auto-hints" setting that should highlight the undo button in red once you're in an unwinnable state: https://uncrossy.com/settings.html
That option shows up in the Puzzle Solved! dialog box.
#Uncrossy Jan 8 in 18 moves, in 2:23, no hints, no resets, no undos, auto-hints on ⬅⬅⬅⬆⬅⬅⬆⬇⬆⬇⬅⬇⬆⬇⬅⬇ https://uncrossy.com
Same here. On yesterday's puzzle I got into an unwinnable state on my very first move, using the topmost word (spoiler: I made "hut" instead of "out").
lmao same. actually a really cool fun/concept it's definitely wordle popularity caliber, but once i got to the last 3 words and ended up in this scenario and the hint button said that i was like -_- owned.
not sure what the right game experience would be for that. a notif that says "You can still solve more words but you'll never solve them all!" doesn't quite work here, because it's sort of saying "there's only one _right_ way to win, but good luck figuring out the right order". Still, it would be better than me finding that out at the very end.
it would probably be pretty important to design levels so that the unwinnable states can't happen early in the game, but it's getting a little abstract to think about at this point. sort of brings me back to that unblock it game from the old ipod touch days.
maybe a percentage chance of solving puzzle tracker that updates a bit randomly slow so you don't necessarily know right away that you made a mistake, although it would have to be a bit weird, for example when you start you are not at 100% of solving puzzle.
There could be a 2 player game where whoever gets stuck (or empty board) loses.
Yes it's annoying, but that's part of the puzzle right?
This was really fun. I actually went back to the beginning of the archive to find more puzzles to solve. I think it would be nice if when you have solved one of the historical puzzles it would give you a button to click onto the next day's puzzle.
I actually never came across the situation where my grid was unsolvable. As I was playing I was wondering what kind of algorithm you used to come up with grids that have no possible dead ends. Since this is clearly an issue, I wonder if it's possible to come up with puzzles that always have a solution similar to how the algorithm for creating Sudoku puzzles always guarantees it can be solved without guessing.
There's a dead end in today's puzzle, if you make TIGHT instead of RIGHT, and the hint will tell you you're at a dead end.
Love the idea. It's fun to play.
I didn't like that it's possible to leave the underlying word unchanged by dragging the word containing a letter twice to the second occurence.
It’s stuck at “Loading…” on both Brave with no extensions and Safari with adblockers. It was only after I tried it on Firefox without anything that I realised the cookie banner is the issue.
In Firefox with ublock origin, I got in the console:
"Loading failed for the <script> with source “https://uncrossy.com/js/cookieManager.js”." looks like a false positive in the ad blocker. Allow-listing the site fixed it.
Same on brave when block is enabled. I think it's because of the cookie banner.
Fun! On first thought, I'd prefer knowing when I'm in an unwinnable state instead of having to keep clicking the hint button.
Also, the site worked for me in Chrome but doesn't work in Firefox (145.0.2). Do `window.cookieManager = ...` (or even `var cookieManager = ...`) instead of `const cookieManager = ...`. This goes for all variables in the global lexical scope you intend to share across source files.
It doesn't work when uBlock Origin blocks uncrossy.com/js/cookieManager.js?v=7
Oh, that's it. My bad!
Worked for me in Firefox 146.0.1.
Ditto.
The website says "I basically need to design puzzles in reverse and have built a set of tools to help me with that." Does that mean that each day's puzzle is essentially built by hand? It seems like an interesting and non-trivial search problem to automatically generate puzzles, given a dictionary.
Uncrossy creator here. Yes, I design each by hand using a highly custom editor that I built to help with many of the common challenges. People keep asking me if it's a job that AI could do, but I think it would be really hard to get right. There's a lot of craft in building the shape of the puzzle and picking words that allow as many spaces as possible to get filled in.
That's super cool! Thanks for explaining. Just to clarify myself - when I was thinking of automation I was not thinking of AI, or at least not how its popularly understood these days. I was thinking more like how 2D mazes can be automatically made by non-AI methods. Uncrossy would need a kind of constraint solver that can navigate words lexically and the spatial intersection of words (and their possible slides). I'm certainly not telling you anything you don't know. There are some algorithms for these (general) kinds of constraint satisfaction problems and they long pre-date "AI", but I think the application to uncrossy and refinement would be quite challenging.
Very nice. Easy to accidentally cheat, however. Shift a word to an invalid position, but right click instead of letting the mouse up event fire. Then shift the word back to the original position: win!
The finishing move on Jan 1st was gorgeous, nice work!
Okay, amazing. (https://uncrossy.com/index.html?p=2026-01-01)
Yes, it was a complete wow moment.
thanks for your comments, got me to play that one, and it's great. very well done!
This is really, really cool. I think telling me how many moves I have to go back in the hint was absolutely a must-do, and shouldn't cost two hints... Second-guessing every single move I made would be insane, but knowing I had to go back seven, and pick something different than the last thing I restored, that worked fine.
It's easy to assume making a word disappear is always the right choice, but you forget it changes the word it leaves behind as well. Very clever.
It does have the same quirk Wordle had that bugged me: Treating browser storage as useful in our multidevice world.
> It's easy to assume making a word disappear is always the right choice, but you forget it changes the word it leaves behind as well. Very clever.
It's also easy to make the opposite assumption, that the goal is to change the other word. I initially felt weird about changing from a letter at position 3 to the same letter at position 1, but eventually realised that the goal is just to slide the word around, not necessarily to make a new word.
Fantastic work, very fun ! I actually only ran into the dead end scenario right until the last few words so not a frustrating first experience. But reading other comments maybe a setting to prevent the player to take a route that ends up unsolvable would be great. Kinda like the "Normal" and "Expert" Modes in worldle
I was hoping this was like the game Crossy Road, but your goal was to throw yourself into traffic
Nicely done! This is an excellent experience, both visually and gameplay wise.
The only thing that felt a bit weird was being able to change the word to the same word, if the slid word had the same letter twice.
Also OMG I've just read your bio - I saw and wondered what someone more from the software side (rather than philsophy) would say: https://open.substack.com/pub/mcauldronism/p/the-maintenance...
Seems like the same idea as treating your LLM as a compiler.
When you write code in Rust and user your compiler to produce an x86 binary, you don't maintain the binary. If you want to make a change, you toss away the old binary, change your Rust code, and recompile.
That concept has been floating around a few places recently. It is wishful thinking. Your userbase won't accept constant small changes in UX, functionality, or the same bugs returning every time you update the production code. And that is what you will get if you run a non-deterministic compiler that hallucinates.
I love this!
Haven't yet really tried the full level but really liked the tutorial, and the quality of the build
update - wow the actual level is better ;)
Really fun. The undo/redo functionality is much appreciated.
I love games where the rules can be understood in seconds
Nice concept! A lot of modal spam could probably reduced, very distracting.
Tutorial was pretty confusing to me. I formed "HIM" and it rejected it, no explanation why.
In tutorial, it did not accept HIM as a solution, because it wanted DIM. Kinda confusing.
You have to form a new word. It starts as HIM.
Was a bit disappointed when I almost "solved" it but couldn't solve the last 2 words, finally clicked the hint and it told me to undo 12 times.. would have preferred if there was a warning earlier.
Oof, visiting the settings page means you lose your place in the game.
Oh that was way more fun than I expected! I feel I shall be playing this regularly!
I love it. My bets are on that this idea will be stolen and turned into some micro-transaction plagued app on store before the end of the week.
Cute. Lots of "-ed" matching tho
Cool game!