Personally, I usually try something before going to docs. Sometimes I exhaust all ideas I have before going to the docs. But I never just do random shit until I’ve tried everything that makes sense, read the docs, and asked around the internet (maybe try random stuff while waiting for replies online)

[–] [email protected] 1 points 1 year ago

I was mostly kidding, though it depends on the problem itself - if I need an explanation for a function argument, no point testing shit if the docs answer it in 15 seconds. If it's something more solution-y, I might do some testing before consulting the papers