A Short Note on Automated Testing
I think people can over test and under test.
Take a simple example of testing a function that takes an integer and expects the integer to be between 0 and 5. You can certainly test all the numbers from 0 to 5, and maybe test -1 and 6 to see what happens if you pass a bad number. Maybe you would even want to test the biggest and smallest integer your languages supports. And just maybe, you would want to test a few random numbers. But you don't really have to test every integer or even every number between -100 and 100. It might be a good way to say you have lots of tests, but it isn't really helping.
At the same time, just testing 0-5 isn't enough, in the same way you can go over-board, you can go under-board.