Feedback and Questions

Hi Andrew,
Your last update to the Sudoku Solver does the trick: Weekly Unsolvable #675 is no longer unsolvable :)

And: I can confirm Jacob's report (see below).
Please load Daily Level 3 from today (12.9.) and step through.
After setting last candidates 3 in D3 and E4, candidate 4 in D6 gets incorrectly highlighted. Candidate 3 for removal in D6 gets not highlighted.

Thanks again for your great work.

I would also like the "Hidden Singles" and "Pairs Naked + Hidden" options.

+-----------------+-----------------+-----------------+
| 3489  26  2346  | 4589 458  7     | 2389 56   1     |
| 189   7   16    | 1589 2    3     | 89   4    56    |
| 13489 12  5     | 6    148  489   | 7    28   2389  |
+-----------------+-----------------+-----------------+
| 6     5   8     | 149  3    249   | 249  12   7     |
| 7     4   12    | 1589 158  6     | 289  3    2589  |
| 13    9   123   | 7    1458 2458  | 6    1258 2458  |
+-----------------+-----------------+-----------------+
| 14    16  7     | 3    9    48    | 5    268  2468  |
| 5     8   46    | 2    7    1     | 34   9    346   |
| 2     3   9     | 458  6    458   | 1    7    48    |
+-----------------+-----------------+-----------------+

If 5R2C9 then

+--------------+-----------------+----------------+
| 489 2   3    | 4589 458  7     | 89  6    1     |
| 89  7   6    | 189  2    3     | 89  4    5     |
| 489 1   5    | 6    48   489   | 7   28   2389  |
+--------------+-----------------+----------------+
| 6   5   8    | 149  3    249   | 249 12   7     |
| 7   4   12   | 1589 158  6     | 289 3    289   |
| 3   9   12   | 7    1458 2458  | 6   1258 248   |
+--------------+-----------------+----------------+
| 1   6   7    | 3    9    48    | 5   28   248   |
| 5   8   4    | 2    7    1     | 3   9    6     |
| 2   3   9    | 458  6    458   | 1   7    48    |
+--------------+-----------------+----------------+

Hidden Singles off & Pointing Pairs on -> 2R45C7 ( ... 2R4C7 )

( -2R4C8 -2R6C8 -2R5C9 -2R6C9 1R4C8
5R6C8 -5R6C6 -48R9C6 5R9C6 -5R9C4
[48]R9C9=[
4R9C9 8R9C4 4R7C6 8R6C9 |
8R9C9 4R9C4 8R7C6 4R6C9 ]
=2R6C6 -2R4C6 2R4C7 )

+--------------+-----------------+--------------+
| 489 2   3    | 4589 458  7     | 89  6   1    |
| 89  7   6    | 189  2    3     | 89  4   5    |
| 489 1   5    | 6    48   489   | 7   28  2389 |
+--------------+-----------------+--------------+
| 6   5   8    | 149  3    249   | 249 1   7    |
| 7   4   12   | 1589 158  6     | 289 3   89   |
| 3   9   12   | 7    1458 248   | 6   5   48   |
+--------------+-----------------+--------------+
| 1   6   7    | 3    9    48    | 5   28  248  |
| 5   8   4    | 2    7    1     | 3   9   6    |
| 2   3   9    | 48   6    5     | 1   7   48   |
+--------------+-----------------+--------------+

Hidden Singles on -> 4R4C7

+--------------+-----------------+--------------+
| 489 2   3    | 4589 458  7     | 89  6   1    |
| 89  7   6    | 1    2    3     | 89  4   5    |
| 489 1   5    | 6    48   489   | 7   28  3    |
+--------------+-----------------+--------------+
| 6   5   8    | 149  3    249   | 4   12  7    |
| 7   4   12   | 1589 158  6     | 289 3   289  |
| 3   9   12   | 7    1458 2458  | 6   5   248  |
+--------------+-----------------+--------------+
| 1   6   7    | 3    9    48    | 5   28  248  |
| 5   8   4    | 2    7    1     | 3   9   6    |
| 2   3   9    | 458  6    458   | 1   7   48   |
+--------------+-----------------+--------------+

(Daily Sudoku 12.08.2025 Level 4) https://www.sudokuwiki.org/sudoku.htm?bd=...4789..2.7.6.............5.6..1..2.73...18.9..7..5.3.............5.4.1..5246...

After the update (2.49 --> 2.50) the solver sometimes incorrectly highlights in yellow the wrong numbers, that have been eliminated by the newly found hidden singles.

Btw, having the strategies "Pointing Pairs", "Box/Line Reduction" and "Triples Naked + Hidden" as optional is a nice addition! What stopped you from also making "Hidden Singles" and "Pairs Naked + Hidden" optional? :)

The Solver is still not working for the Jigsaw. When loaded into the Solver the numbers appear to be OK, but the grid changes.

https://www.sudokuwiki.org/jigsaw.aspx?shape=16&bd=...3.2..47.8......3..64......5......6..2.4..5......3..1...2..........4.84..517...

For the past 2 nights, Dec. 4 & 5, 2025, the Jigsaw solver shows the given clues in the correct positions row & column, but the pattern in the Solver is different from the pattern in the Puzzle. If I solve a cell & check with the Solver, same problem, wrong pattern from the Puzzle yielding 0 Solution Count

This summer, a change was made to refuse to process any sudokus with less than 17 givens. This has made your tool significantly less helpful for variant sudokus, such as Thermos, Killers where some regions have no sums, or Just One sudokus where only one digit can be deduced with confidence.

Would you consider changing the 17 clue requirement from an error to a warning, and still let the existing logic do what it can?

Playing on str8ts.com with my Android phone the browser annoyingly tries to select the text on the number-buttons when i use them.
Could be an easy fix, to disable text selection for the buttons via CSS:
.numberpad {
(...)
-webkit-user-select: none;
user-select: none;
}

Thank you for providing lots of free puzzles!
Greetings
Leo

The new Killer entry system with the number selection is excellent!

What a great site. I came just looking for new patterns for solving beyond what I had done for years, and I ended with a perennially open tab in my browser!

I'm truly impressed by the complexity of some of these solving methods, yet you have code to pull them off as well! I was in a long hiatus from Sudoku for many years, and then one day prompted an AI chat with a NYT sudoku I was stuck on... It UTTERLY failed to comprehend Sudoku at all. Tried, but failed. Then I fell in the rabbit hole: Can I write some front end to solve and create puzzles? How do you break down the steps we take in our heads and while solving? ...and so on. Not sure I came out on top of it all, but i did get somewhere: http://dice.phillipwildhirt.com/sudoku

Anyways, found your site and now the world has opened up! Just makes me want to start all over and try again! Thanks for your work!

Hi / I’m not sure if it’s me, but I have noticed the following issue since using the new Version 2.47.1 of the suduko solver on my iPad :-

If you select a cell in the solver to edit the candidates (ie touch the cell on the iPad screen) and then realise it’s not the cell you wanted , you used to be able to immediately select another cell. With the new software if you select a second cell without hitting the return key first, it blanks (hides) the contents of the second cell selected.

You can make the cell candidates visible again by toggling the candidates shown tick box off and on, but it wont allow you to edit the candidates.

The only option I can find is to delete the ipad page and reload the solver from the daily page, thus losing all my edits in the solver.

I’ve tried removing the solver from memory and rebooting but to no avail.

Tested on 2 separate iPads.

Regards

Jon