3 ms·
In the step of √2 = a/b you can get away without asking for relatively prime, using GCD or this descent by asking for least a for which this holds. Then when yo
by IsTom 2y ago
In the step of √2 = a/b you can get away without asking for relatively prime, using GCD or this descent by asking for least a for which this holds. Then when you get a/b = 2c/2d = c/d, but c would be less than a, hence contradiction.
Couldn't you always do this instead of doing this infinite descent? If relation is well-founded then there will be a least element.