Jeff wrote:When you let r3c8=1, a contradiction (an error) was found. So, you backtracked and let r3c8=5. This solved the puzzle and you should have stopped there. Letting r3c8=7 was redundant.
Jeff wrote:The identification of an implication chain is T&E if you choose to use a T&E method to identify it, like the one you have used. There exist some advanced techniques such as bilocation/bivalue plot and advanced colouring table, which would remove the elements of trial and error from the identification process. Such identification process is termed 'pattern recognition' and is non-T&E.
Well, I was JUST ABOUT to be convinced.
But as u see the contradiction acts only as a STOP signal (No consequences) because I chose it to be so, I disagree on what you said that backtarcking after finding a contradiction then makes it T&E because then every technique that spots a contradiction would constitute T&E. I can easily choose the step before reaching the contradiction as the stop signal & avoid all of this, that can't be right
.
Now the bit to me that was more convincing was the issue about "recognition patterns", but there are some issues hinged on that statement.
All pattern recognition techniques have to look for the pattern. LOOKING=TRIAL, the machine or solver goes through many dead ends before spitting out the pattern. what you describe as recognition patter is the end result of tedious search with trial in each step. The humans do the same to alesser extent but they become more machine-like as they use pencil-markings & solve it systematically.
Now I concede that the result came before the "How I did it ?", but knowing that the a demonstrable easily constructed non-T&E pattern does exist anyway for each step (which I did show), then I can bulid an argument from there stating that by the way " This non-T&E technique proves it". T&E can't do that.
Now if I can present EVERY result I have as consequences of a double implication chain or an xy-chain, then why shouldn't I use it in a helper program & in this forum ???
I still need more convincing, what about you
?