https://www.steamgifts.com/discussion/IJJwy/rosimincs-2025-nonogram-cafe-weeks-8-11-active-2-bonuses/search?page=8#EWmgUQQ

something definitely happened

I'm still alive I swear!! :D
I have started a new project at work and it's been keeping me busy and tired haha, my wife and I have been playing Farm Together 2 on evenings to relax and I was procrastinating on making the new puzzles ^^

Thanks a lot, featured on the main post! :)

Solved, thanks for the nonogram!

I see nothing wrong with this one, only one solution, no need to guess.

Interesting thought experiment :) I personally am in the camp of do-not-ask-me-to-guess-a-spot-even-if-the-solution-ends-up-unique.

Sure, small puzzles like your example are not an issue, but the complexity can scale fast. It's also why I grudgingly do not like puzzles such as Lexica because for bigger puzzles, you either need a huge capacity for thinking ahead (not me), or you'd have to do an educated guess on a spot and work from there (defeats the purpose of a logic puzzle for me). I like the concept though, hence "grudgingly".

I know that when I was doing research on valid types of nonograms I encountered those, and saw different opinions on them as well.
For me personally, I always tend to abandon a puzzle when there is some type of backtracking/guessing/forward-visualizing to them because I don't have a very good visual memory. I wouldn't say that this kind of puzzle is invalid though, just that they're a bit more advanced solving-wise.

Now when it comes to the algorithm I use to create and validate nonograms, it would consider your example as invalid, because I make conclusions out of checking individual rows/columns at a time.
For example, if a row has 5 cells and the hint is 2 1, I get all possible combinations:
<X X _ X _> <X X _ _ X> <_ X X _ X>
Then from those combinations I evaluate if a certain position always has X or _ and mark it
With that logic, I conclude that the 2nd cell is filled.
Using that logic only, and simplifying the possibilities according to markings, I can find the solution or determine if a puzzle is 'invalid'.
However, a valid puzzle like yours will be flagged as invalid because the solving algorithm I used cannot think ahead.

Doing it that way also ensures that puzzles I generate are not overly complicated, so that they can always be accessible to more casual solvers (granted, the 25x25 grids can be quite daunting for newcomers but still ^^)

I like the examples you gave because they simplify the concept a lot, I always have trouble visualizing those kinds of problems haha
And the logic behind them is quite cool, I think that those valid-with-guessing puzzles can bring up the challenge even higher (too high for me though!)

i think i realized the problem, i may be moving my mouse a couple pixels each time i right click? even one pixel movement deactivates the x, like it thinks i want to hold and drag... but placing multiple x in a row is a rare input for me, so i think the hold and drag should only be enabled for lmb?