My worst nightmare:

it's not the code what's wrong during testing.

It's the unit tests themselves

Time to write some tests for the tests
Then some tests for the tests for the tests
Then some tests for the tests for the tests for the tests
Then some tests for the tests for the tests for the tests for the tests
Then some tests for the tests for the tests for t...*blink*

Error: human.brain.OutOfMemoryException : stack overflow

My moment of rare happiness: seeing a test failing because of a real problem and not the fucking test itself.

@redundandundant totally understand that one, hardly ever happens to me at the minute; project I'm working on is too new, structure keeps changing.

what's worse is that both the code and the tests could be wrong but the tests could still pass.

We have a separate test team to maintain our CI machines and write the tests. Don't get me wrong, I'm very thankful for their existence but the day before the deployment we had 50% of the tests failing and we had to stay up, and our test team identified that the tests are broken. But you know, it takes a Jenkins machine 6 hours to complete a full test 😁

I'm not at all salty, it could've been me who wrote the buggy code AND furthermore pushed that to the repo!

Okay I would bang my head to a wall if that happens, but before that I will bang the head of the person who wrote those tests

Not only that my code was wrong, but also the unit tests returned the expected results. Several days went by till I recognized my mistakes...

Never code when being tired!

Tests are code too. It’s basically a “yo dawg, I heard you like code, so I put some code in your code”

No replacement for some good old fashioned human testing.

@Brolls this made me think what the test suite for a human would look like...

You know we'll need this for AI in the future