This post not CC'd by email On 1 Dec 1999 21:20:15 +0100, Bertram Felgenhauer <email@example.com> wrote:
>> Is it soluble by Boolean algebra or matrix methods? > >What I usually do with these is to imagine "1"s in the places where I >put "*"s and then to look for rectangles with corners like > >1 - 0 1 - 0 1 - 1 1 - 1 1 - 1 1 - 1 >| | --> | | , | | --> | | , | | --> | | ; > - 1 0 - 1 - 1 1 - 1 - 0 0 - 0 > >basically I use the fact that rectangles with exactly one corner >containing a "0" would lead to a contradiction, so I simply avoid them. > >have fun...
Thids method is so neat. Thank you. At first I thought it was like parity checking but the rules are different. Still it could be done with logic gates.
Thanks for a most thought provoking answer.
Are there other rules?
For instance is there a rule that indicates when sufficient 0s and 1s have been extracted from the clues? Optimising the solution process would require desisting from the clue sifting stage as soon as sufficient basic 0s and 1s had been extracted.