Comment by staunton

Also a lot of errors would be called "typos", not errors. Such as some edge cases missing in the theorem statement which technically makes the theorem false. As long as there's a similar theorem in the same spirit that can be proven, that's what the original was all along.