Some puzzles may require very complex logic to determine that they have no solution or have many solutions.
Are there discussions on this subject?
dobrichev wrote:Are there discussions on this subject?
dobrichev wrote:Some puzzles may require very complex logic to determine that they have no solution ...
Are there discussions on this subject?
dobrichev wrote:For determining multiple solutions, counting for at least 8 different given digits and looking for no two empty rows in a band can be considered most basic techniques, "singles".
dobrichev wrote:Is Denis Berthier's rating sensitive to puzzle validity?
dobrichev wrote:For determining multiple solutions, counting for at least 8 different given digits and looking for no two empty rows in a band can be considered most basic techniques, "singles".
blue wrote:I almost mentioned this earlier, but didn't know if it was relevant: there are puzzles with fewer than 8 digits and/or with 2 or 3 empty rows in a band, that have no solution, rather than multiple solutions.
blue wrote:Would the goal be to determine the final disposition, or only to determine whether it does/doesn't have a unique solution ?
dobrichev wrote:The multiple-solution case shouldn't be a stopper if
- you take care when using resolution theories that assume uniqueness of the solution. Most trivial patch is to disable these theories for this particular case.
dobrichev wrote:Is your rating system materialized in an open source software?
dobrichev wrote:Is it possible (after minor modifications) pencilmarks to be used as input instead of ordinary puzzle?
denis_berthier wrote:2) a contradiction (together with the path leading to it - which is a constructive proof of the contradiction); in PBCS, I've given an example of this case, showing that proving a contradiction can be as hard as finding a solution; (indeed, I had given examples much before, but the forums have disappeared); ...
m_b_metcalf wrote:denis_berthier wrote:2) a contradiction (together with the path leading to it - which is a constructive proof of the contradiction); in PBCS, I've given an example of this case, showing that proving a contradiction can be as hard as finding a solution; (indeed, I had given examples much before, but the forums have disappeared); ...
Denis,
If the Programmers' Forum is the one you meant, you'll find an archived copy here.