I think people trusted that computers could safely do arithmetic operations. (Though later chip problems have shown this trust needs to be earned!) But a proof based on running a long computer program is hard to check. I think the problem people had with the proof was that they didn't know how to verify that the computer did what it had to do. For example, the Kepler conjecture was presumably proved in 1998 by Thomas Hales with a complicated proof that partly relied on a similarly long computer program. I think that, after 10 years, a committee of mathematicians was able to give its imprimatur to the non-computer part of the proof, but gave up on trying to verify that the computer did what it was supposed to do. --Dan On 2013-04-29, at 10:37 AM, Robert Baillie wrote:
I never saw what the big deal was about this computer proof.
For many years before 1976, computers had been the only entities that were capable of verifying the primality of large Mersenne numbers.
Bob Baillie