[docs] parameters var-name usage about rbs HOT 2 CLOSED

Thanks for pointing out this and starting the discussion.

1 concern is I dont know what it means to me.

It doesn't mean anything for keyword arguments. The parameter names for positional arguments make some sense I think. It can be used for documentation. We often want to refer to the arguments in the documentation.

Your concern makes sense because keyword arguments can be referred safely by their keyword names.

2 concern is code drift.

There is no one-to-one correspondence between the parameter names in method definitions and the names in RBS. We can give more precise/better definitions in RBS, like the following example.

# def foo: (foo: String, bar: Integer) -> void
#        | (baz: String) -> String
#        | (**String others) -> untyped
def foo(**opts)
end

3 concern is correctness.

They are the same.

I'm thinking of just removing var names from keyword arguments. I don't think it causes serious problems, but agree that we haven't found good usage yet.

from rbs.

Hello @soutaro,

Thank you for your response. I think the biggest question I had was why and that was answer. The main reason is for documentation. The name of the variable can be documented but it has no impact on the code or signature. This would help in a a code hinting system where the comments would not need to be parsed. Your example of def foo is a very informative as well.

Should I leave this issue open to track removing keyword argument var names?

Thank you once again for your time,
Dictated but not reviewed,
Becker

from rbs.