How often and in which cases do you write comments in your code?

- When I need to explain a weird design decision
- When the code itself isn't straight forward and needs to be explained
- todos

@alexbrooklyn yeah explaining weird design decisions so I will understand them in 6 months time
⊙﹏⊙

Usually only when something is not immediately obvious.

Well, usually for generating api documentation.

- consumed api docs over each method [if docs change - methods will too]. I find this more useful than looking for docs online every time
- the WHYs that aren't and cannot be made obvious
- todos
- copyright
- if the code is to be shared - interfaces' methods.

Most of times as a way to get over writers block. Pseudo code, and todos.

Later I remove most of them.

Docs for future reference, todos, and when I'm bored/frustrated.
I usually mantain conversations with myself through time through comments, like:
@highlight
//maybe use that thing YEAH GOOD IDEA not anymore, that's gone

Never

Not even a joke

Anything that isn't immediately obvious, or that will help a future dev. Also, explanations of why, and sometimes section markers.

When there is a chance not doing so will blow someone up in the future. I try to be empathic about the people who will be maintaining my code and avoid being a justifying case for, "that last dev's code was complete shit."

I comment pretty much everything, which might be excessive, but I want to make sure that the next person isn't left with too much spaghetti on their plate.

I keep my comments short and to the point, mainly to explain the algorithms as I write them in pieces.