I've received a lot of interesting comments and questions from Sudoku fans over the last few years and this page
is where I try to answer them. I'm also directing Str8ts feedback here. Please feel free to drop me a note on the side of the page. Or you can email me directly at andrew@str8ts.com.
The first board is classified as unsolvable, however, the second board, with more clues, is solvable according to the solver. All strategies are used. Is this intended, because there is not a unique solution or is this a bug? Here are the boards: 100000000020000000003000000000400000000050000000006000000000700000000080000000009 145000000726000000893000000000412000000753000000896000000000712000000483000000569 Regards Martin
Andrew Stuart writes:
The solver only works logically if there is a unique solution since there is no logical way to choose one over the other if there are multiple solutions. Solution Count is there to check you have a valid puzzle.
Tuesday 21-May-2024
... by: Mary, Canada
I just tried to restart the Daily Kakuro and the restart button doesn't work. Love your puzzles!!! - Thanks. Mary
Andrew Stuart writes:
Fixed! Found the problem. Refresh until you see v2.03. Thanks for the report
Sunday 19-May-2024
... by: BdR, Netherlands
A local newspaper has a prize Kenken puzzle every week, and I was unable to solve it by hand. I wanted to check with this solver because I feel like I'm overlooking some method or approach (to be clear, I wasn't going to cheat and send in an automated-solution). However, when I press "step" after 2 steps it gives the error message "Some cages are larger than 6 cells. 7 is not legal for a valid KenKen 6x6 puzzle" even though it should be solvable. LINK
In the same paper from previous week, this Kenken puzzle is also unsolvable somehow, the auto-solver finds the same top-left digits but I don't know which step to do next LINK cheers, Bas
Andrew Stuart writes:
Yes I don't know what to do with cages that > 6 cells. I consider them a different type of puzzles. I might look at the code and see if I can adapt it. Does make the puzzle a lot more difficult
The KenKen solver lags behind the Sudoku solver when it comes to the most difficult puzzles. Still some strategies to be discovered. The problem is too many candidates at the start. Very few Sudoku strategies work with so many
Saturday 18-May-2024
... by: Kaz, Netherlands
i am unable to edit candidates in the color sudoku, pressing enter does nothing, pressing outside of the cell, returns everything as it was before (other cells too)
Andrew Stuart writes:
Not sure what you are experiencing. Do the other solvers work? Colour Sudoku works for me. Maybe uncheck the "auto-clear" tickbox? Write with an email so I can contact you directly.
Friday 17-May-2024
... by: blortch1, Hawaii
I suggest moving APE into the Extreme Strategies section. APE appears to require a computer - unless one has an unlimited amount of time in their life - to detect where it could be used.
Andrew Stuart writes:
Might swap with Grouped X-Cycles but then no APE might appear. Will have to do quite a bit of testing but I understand your assessment
Tuesday 16-Apr-2024
... by: Aubrey de Grey, United States
I'm finding that I can't turn off Death Blossom - puzzles use it even when it's unticked. I've seen this three times now - I don't think it's a refresh issue.
Andrew Stuart writes:
Yes this was a feature rather than a bug. When I look for the previous strategy "Almost Locked Sets" I get a "Death Blossom" result for free. However I can see that is confusing with the checkboxes so I now null the result if DB is checked off
Saturday 6-Apr-2024
... by: ,
I opened the SudokuSolver app on my Android tablet. When I chose Daily Puzzles, the app just hung at downloading today's puzzles (for April 5). I couldn't back out; the only option was to close the app. I tried a few times. Same result. I restarted the tablet. Same result. I went to the Google Play Store in case there was a new version. Nope. I am up to date.
Andrew Stuart writes:
Working for me, just tested now. Version 1.85. Can you get back in contact with an email address. Try and figure this out
Thursday 28-Mar-2024
... by: Dr John Collins, Australia
Whenever I enter a new kenken I never get a clear box for the solution. It always leaves a solution so I have to remove each digit in the solution for each of the 36 entries.
Andrew Stuart writes:
I will add a button. Good idea
Thursday 28-Mar-2024
... by: Link, USA
I'm writing a solver in Python. I could benefit from a feature on your site that downloads (or displays a copy-able string) of the 81 character array from your puzzles. This would help me greatly in testing. Thanks. Link Reynolds
Andrew Stuart writes:
If you use the [Email this board] button you get two strings, the 81 char and the longer one with candidates Or do you want something more API-ish?
Saturday 25-May-2024
... by: Stephen Fairchild, United Kingdom
Load Sudoku: CLICK TO LOADWith the supplied broken 17 clue grid both 5 and 9 are highlighted for F9 then the 9 is promoted without regard to the 5.
Situation I’m certain should only occur if the puzzle is broken (multiple solutions)
Thursday 23-May-2024
... by: Martin, Germany
Load Sudoku: CLICK TO LOADHere are the boards:
100000000020000000003000000000400000000050000000006000000000700000000080000000009
145000000726000000893000000000412000000753000000896000000000712000000483000000569
Regards
Martin
Tuesday 21-May-2024
... by: Mary, Canada
Love your puzzles!!! - Thanks. Mary
Sunday 19-May-2024
... by: BdR, Netherlands
LINK
In the same paper from previous week, this Kenken puzzle is also unsolvable somehow, the auto-solver finds the same top-left digits but I don't know which step to do next LINK
cheers,
Bas
The KenKen solver lags behind the Sudoku solver when it comes to the most difficult puzzles. Still some strategies to be discovered. The problem is too many candidates at the start. Very few Sudoku strategies work with so many
Saturday 18-May-2024
... by: Kaz, Netherlands
Friday 17-May-2024
... by: blortch1, Hawaii
Tuesday 16-Apr-2024
... by: Aubrey de Grey, United States
Saturday 6-Apr-2024
... by: ,
Thursday 28-Mar-2024
... by: Dr John Collins, Australia
Thursday 28-Mar-2024
... by: Link, USA
Thanks. Link Reynolds
Or do you want something more API-ish?