StrmCkr wrote:so y woudnt it be possible to identify a rectagle buried in an intial set of numbers and determine a way of deducing which of them is the N corrorner to remove. instead of having to slowly reduce numbers to express the rectangle in entirety as singlualry expressed then solve it.
Sure, it could possible to do that. But then you have to find a binding proof for your deduction, which you haven't showed here yet. You just mention two rectangles (that are very far from becoming deadly patterns) and based on that throw out some solved cells. You have to show that if your solved cells are not true then a deadly pattern (or other contradiction) occurs. Until then your deductions are invalid. And remember that it's not a question of "which corner", it's "which corner
s" will not have the deadly candidates. For both of your rectangles it's very possible that none of the corners will have any of the candidates you mentioned.
Also, in the other example (which I assume should say r8c3=4 and r8c4=1), even if you did determine that those cells cannot hold the candidates of your two rectangles, what happened to the other possible candidates in those cells?
RW