# Kakuro Research Paper

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them.

The difficulty settings are a little flimsy, but in general a hard puzzle will be more difficult than an easy puzzle.

The easies have a tighter grid and simpler logic deductions, while the mediums have a slightly more sparse grid and tougher deductions, and the hards have the most wide open grid while keeping the deductions the same as medium.

Although maybe that's just me being bad at kakuros. I'd also love to hear a good algorithm for detecting naked triples!

As most of your puzzles have obvious surface sums, it's strange that you don't use them.

It's several orders of magnitude greater than simple Latin Square cases such as Sudoku/Skyscraper.

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] 0[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] 30[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] 45[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] 35[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] 37[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] 6[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] 21[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] 24[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] 0[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] 35[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] 17[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] 42[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] 28[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] 0[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] 16 - - - - - - - 0 - - - - 0 - - - - - - - - 4 - - - - 0 - - - - - 30 - - - - 35 - - 0 - - - - - - - 17 - - - - - 0[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] 17[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] 29 - - - - - - 37 - - - - 0 - - - 11 - - - - - 19 - - - 0 - - - - - - 45 - - - - - - 0[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] 27 - - - - 16 - - - - - 4[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] 11[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] 0 - - - - 7 - - - - - 30 - - 0 - - 13[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] 0 - - - 15 - - - - - 0 - - - 0 - - - - - - - 29[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] 30[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] 0 - - - 17[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] 24 - - - 4[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] 0 - - - 0[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] 3[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] 11 - - - - - - - 0 - - - 0 - - - - - 21 - - - 11[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] 45 - - 0 - - 35 - - - - - 44 - - - - 0[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] 29[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] 41 - - - - - 10 - - - - 6[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] 0 - - - - - - 39 - - - - - - 0 - - - 32 - - - - - 35 - - - 0 - - - - 15 - - - - - - 22[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] 10[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] 0 - - - - - 22 - - - - - - - 0 - - 17 - - - - 15 - - - - - 0 - - - - 0 - - - - - - - - 0 - - - - 0 - - - - - - - 0[[

The chances of any random grid producing a well-formed puzzle (ie.

I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.

The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.

It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.

Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.

||

The chances of any random grid producing a well-formed puzzle (ie.I liked the idea of every solve iteration having at least one solution, so it wouldn't waste its time with totally unsolvable puzzles.The solver is coded so that if a puzzle has multiple solutions it will fail to find any of them.It'd also be nice to have the puzzles named some way, so that we can refer to them without having to post them. to what Conceptis uses for their kakuro books, at least, their puzzles and mine have lots of similar attributes.Including the fact their hards and my hards aren't very hard for a seasoned solver. I am also trying to develop Kakuro generating software, and am also familiar with Conceptis puzzles. PS: You must have also noticed, as I have, how very little information about puzzle-generation there is available in the public domain.The generator used for the daily puzzles can be set to produce much harder puzzles, but they can take a time to generate, and it's a Java program, so it's not available to the public.I also want to keep the daily puzzles' difficulty reasonable, since I personally don't find a ton of enjoyment in incredibly hard puzzles without obvious deductions.This example is puzzle #103: 24 rows 14 cols 0\0 0\0 30\0 45\0 35\0 37\0 6\0 21\0 24\0 0\0 35\0 17\0 42\0 28\0 0\0 16\42 - - - - - - - 0\29 - - - - 0\36 - - - - - - - - 4\19 - - - - 0\35 - - - - - 30\20 - - - - 35\15 - - 0\28 - - - - - - - 17\26 - - - - - 0\0 17\0 29\39 - - - - - - 37\21 - - - - 0\19 - - - 11\22 - - - - - 19\9 - - - 0\32 - - - - - - 45\33 - - - - - - 0\0 27\23 - - - - 16\16 - - - - - 4\0 11\0 0\29 - - - - 7\29 - - - - - 30\3 - - 0\14 - - 13\0 0\7 - - - 15\30 - - - - - 0\20 - - - 0\33 - - - - - - - 29\0 30\0 0\8 - - - 17\0 24\7 - - - 4\0 0\21 - - - 0\0 3\0 11\42 - - - - - - - 0\23 - - - 0\18 - - - - - 21\7 - - - 11\0 45\13 - - 0\4 - - 35\16 - - - - - 44\27 - - - - 0\0 29\0 41\27 - - - - - 10\23 - - - - 6\0 0\39 - - - - - - 39\21 - - - - - - 0\11 - - - 32\24 - - - - - 35\10 - - - 0\30 - - - - 15\33 - - - - - - 22\0 10\0 0\34 - - - - - 22\28 - - - - - - - 0\3 - - 17\22 - - - - 15\33 - - - - - 0\28 - - - - 0\36 - - - - - - - - 0\22 - - - - 0\31 - - - - - - - 0\0 However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

]] However, based on the way top-down Sudoku generators work, I imagine all the generators start by choosing grid size, then pattern of black/white cells ...then (more or less randomly) fill the white grid, then compute the sums in the black cells, and then delete sums one by one as long as unicity is preserved. I use just that method for generating Skyscraper puzzles, which was my first venture into puzzle-creating software. It's got a timed daily puzzle contest and a solver for registered users, but the generator is available for everyone.It can generate puzzles up to 20x20, and it runs entirely in the browser in javascript, so there's no applet or anything to load.

• ###### National Academy of Sciences

PNAS is one of the world's most-cited and comprehensive multidisciplinary scientific journals, publishing more than 3,200 research papers annually.…

• ###### The Dewey blog Mathematical or not?

Jun 20, 2006. So we assigned Sudoku and Kakuro the Dewey number 793.73 Puzzles. The Wikipedia even has an article on the Mathematics of Sudoku.…

• ###### Computer Science for Fun - cs4fn Computational Thinking.

Automata Booklet Build a paper puzzle. explore finite state machines. Kakuro. A fragment of a Kakuro puzzle Logical thinking plus addition.…

• ###### New Kakuro Generator/Site Kakuro - The New Sudoku Players' Forum.

Dec 9, 2013. I've just made live my new Kakuro site. It's got. BTW, how does your generation process work. Sometimes that fails too - I lost a whole swag of research material on Gerechte designs, ie non-standard Sudoku/Skyscraper.…

• ###### Color can make a large, flat kakuro puzzle look a little less.

Color can make a large, flat kakuro puzzle look a little less frightening. A flat layout on larger kakuro puzzles can be confusing. Sudoku Puzzles, Logic Puzzles.…

• ###### Ponder this - IBM Research

IBM Research - Haifa is the largest lab of IBM Research Division outside of the United States. Founded as a small scientific center in 1972, it grew into a major lab that leads the. Work with us. January 2015 Kakuro - easy version.…