LÖVE

milon wrote: ↑Tue Jun 22, 2021 7:31 pm Fun game!

Not sure if it's a bug or a feature, but the web implementation does not enforce a unique solution. I just encountered a game where there were 2 possible solutions. I took a screenshot, but somehow the picture didn't actually save.

I saw it too, but both of them are correct.

darkfrei wrote: ↑Tue Jun 22, 2021 8:06 pm

milon wrote: ↑Tue Jun 22, 2021 7:31 pm Fun game!

Not sure if it's a bug or a feature, but the web implementation does not enforce a unique solution. I just encountered a game where there were 2 possible solutions. I took a screenshot, but somehow the picture didn't actually save.

I saw it too, but both of them are correct.

Right, that's what I'm saying: Sometimes there's more than 1 correct solution. That may be a bug or a feature, depending on the author's intent. It wasn't clear to me which case it was, so I was bringing it to the author's attention.

I thought, enforcing a unique solution wouldn't be good. It calculates a random grid, calculates the column and row numbers, then clears the grid. Downside is two different seed values can give you the same puzzle.

Okay, feature then. Thanks for clarifying.

And I'd expect seed collisions like that would be rare enough to not matter.

Improved user interface and new single solution mode

Nice, which UI lib did you use. Your own or...?

Yes, my own

Just tried out your latest update on the downloaded Love2D version and think I have run into a problem. Not sure if it is a problem or the way it is designed. In short - some of the clues show positive (ie that the line is correct) when it's not - your 'Left' counter shows 0 when there are still incorrectly displayed lines. You may have designed it that way, but I find it confusing. An image attached to show you what I mean.

On another point - I have tried to emulate your NoNo on an iPad using Codea (Included under the same licence as yours), and have made it work. The reality is it's got a similar front end to yours but with totally different code (my coding is far more disorganised than yours). Just thought I'd mention that I use checking for horiontal and vertical lines by creating a string and comparing that to a target string and it works for checking each line. Also a check for reaching the complete solution is just done on the horisontal lines. You can download my version from the Codea website if you want to check it out.

Thanks for publishing this I have really enjoyed adapting it to Codea and I find it a little addictive.

Your NoNo looks good.

For the "Left": yes, it is a little misleading. It is "squares in the solution" - "squares on the grid"
So, it can be negative too. "Left" makes more sense in single solution mode.
I can remove it since there are already column/row indicators

Thought it might be that. But if you look at the image I posted - all the clues on the left say resolved whereas two at the top say unresolved.

In the code I made for Codea the clues only change to resolved with the correct number of tiles in the correct positions. I have a routine which checks each line horizontal and vertical. When you get to the final tile to complete the puzzle I thought you would find it easily because it is at the intersection of one vertical and one horizontal line (ie two unfinished lines).

I may be wrong but I thought that was how they were supposed to work so programmed it into my code. I only found this in your latest love2d download, haven't tested your webpage version.