Forums

Ah!

Sometimes, in a puzzle, you know that if a certain step would yield a solution, then there must be another solution. As each puzzle is supposed to have a single unique solution, you know that that step must be wrong.

I personally tend not to use this information, as it feels like cheating, because I would be using meta-information which I don't consider to be part of the puzzle.

How do others think about this? Is using the fact that each puzzle has a unique solution cheating, or do you consider it to be just another clue?

How about "Within each line, column, and demarcated area, each digit occurs the same number of times, and at least once"?

No amount of "implying" will change the fact that the "solution" does not correspond to the rules.

When "solving" this puzzle you have to conclude that the puzzle is wrong, and then throw one of the rules overboard.

Klein foutje: in het middelste vakje kun je een 6 invullen.

Number: Puzzle #780
Genre: Sudoku
Author: Johan
Appeared at: April 20, 2008

You could pop up a confirmation dialog. One with a "Do not ask this question again." checkbox to prevent annoying those who know what they're doing.

Another option is to add an "undo" for this operation (perhaps replacing the eraser icon).
There could even be a generic "undo" option for all actions. Just being able to undo a single step could prevent a lot of frustration.

Similar to how changing the sort order broke the puzzle list when one has the "solved/unsolved" choice open, the same things happens when you select "next" or one of the other links which show a new page of puzzles.

You get this effect when you are not actually logged in anymore, but an old page which shows you as logged in is loaded from the browser cache.
Just reload the page, and relogin.

It may be better if the server would tell your browser not to cache these pages.

Ah, nice.

I have my doubts about this puzzle (or the solution rather).

I have solved this puzzle twice (at least I think I have), double-checked the results, and yet the "Check" button reports this puzzle as incorrect.

Could someone please confirm that this puzzle can be solved in the eyes of the checker? I don't want to use the "Reveal" button.

I see in the statistics that only one registered user has solved this puzzle, after half a month. Could that be the creator of this puzzle himself?

"Congratulations! You correctly solved the puzzle."
As opposed to "incorrectly solved"?

1. The puzzles where you would use this feature are the ones which need backtracking, which are the ones you wouldn't want to do on paper anyhow. It would be nice if all of the puzzles could be solved without backtracking, but unfortunatly this is not the case.

2. The player is still doing all of the reasoning. A "clear everything of this colour" button or an undo button just tells the applet exactly what to do, only more effeciently than having to click each entry of that colour individually.

3. Having such a feature allows the player to spend more time on the actual puzzle solving, and less on the bookkeeping. And imho, the puzzle solving is what it is about.

4. What stays closer to paper version, a "fast erase" button, or duplicating the tab for each branch you take?

5. What's wrong with with offering tools that take advantage of the medium? Electronic form has its own disadvantages; so what if it has some advantages too?

When I complete a puzzle, and press "Check", sometimes I get the message that the puzzle is incorrect, while it really isn't.
Only after I select "duplicate tab "in my browser (Opera), giving me the completed puzzle in a new tab, and press the "Check" button there, will I get the "Congratulations" message.

If I quickly change several fields in a puzzle, sometimes a field is not updated onscreen. The puzzle logic itself does seem to register the change.