Puzzle Description Suppressed:Click below to view spoilers

#1: Susan Eberhardt (Susaneber) on Jul 1, 2022

And my birthday. I've had a lot of birthdays and my bikini days are far in the past.

It was my mom's birthday, too. She'd have been 107 this year, but she died of a second stroke in 2000.

Comment Suppressed:Click below to view spoilers

Happy Birthday Susan! And Gary, I'm sorry for your loss.

This is definitely CL
In a couple places you have to follow pixel by pixel what can and cannot be a certain color

After LL/ easy CL
The remaining red in c2 can only be in r12
mirrored: the remaining red in c8 can only be in r12
LL

The remaining blue 1s in r13 can only be in c2 & c8 (all other spaces can only be red or white)

Then LL to finish

I tried it again and *was* able to solve it with color logic. Joe, thanks for the encouragement to try it!

Glad to help, Valerie. The reason that the helper won't solve this is because it only does line logic. Color logic necessarily uses information from the crossing row/column.

The helper uses regular expressions to check for overlapping clues. It won't even find 1s in a row of all 1s if it is unsure of *which* 1 a pixel is.

The create puzzle solver is much more capable because it runs on the server and is in C iirc.

The solver is never wrong about LL/CL. If it cannot figure it out it leaves the solvability blank. It used to be able to figure out a lot of moderate lookahead, too, but I think that got lost when it got rolled back with the site loss

Hm. I don't think the solver should have lost anything in the 2016 database crash, because the code for the site stayed intact through that; what was lost was only the database.

I know the solver doesn't do color logic. But I do, so I should have been able to solve that!

Show: Spoilers

Goto next topic

You must register and log in to be able to participate in this discussion.