Copilot. Love it or hate AI, copilot is a useful tool but after using it for a couple of weeks, with intense activity over the last few days I have to say I have mixed feelings about it as a tool.
I think like any other tool, or human for that matter, there are times copilot gets it right, and there are times it really gets it wrong.
To give an example of where it seems to consistently get it wrong, I want to look at testing.
Now when it comes to testing in golang
I am not at my finest. In previous roles I was almost pious about it. "You will write tests, you must cover every line and every branch" - but after an extended stint in a company where writing tests wasn't just frowned upon but was actively discouraged, I found my own attitude to writing them changed. Therefore, until recently I've had an extremely relaxed attitude towards them which has left me learning how to write them all over again.
As I have come to write more and more in golang
, I find myself in a position where I not only need to wr