Re: [External] : Re: Shrinking the C core

[Arthur Miller]

Drew Adams <drew.adams@oracle.com> writes:

> [Moving reply to e-tangents.]
>
>> > I would not object to using keyword arguments for functions
>> > like that -- complex and cumbersome to use, and not used often.
>> 
>> Well, than we pretty much agree 100% about keyword args.
>
> +1.
>
> Except maybe for the "not used often".  Depends on the
> "often" use.  If used often without any keyword args (or
> optional args, for that matter), what difference does it
> make that the function _has_ keyword/optional args?  By
> definition and convention, such args are generally not
> for the most common (the default) use case.

Well yes, of course, it is corect. I think it depends on what "not often
used" referred to. I interepretted it as not often used argument, not
the function itself. Those more often used but possible to omit are of
course usually optional and in front of the optional list, but those
that are used in some more specialized circumstance or by very few are
perhaps good candidates for keyword arguments? I don't know, that is how
I perecieve it.

> But the general point, I think, is that keyword (and
> optional) args turn one function into a family of several,
> and that can be useful.  But the "base" function, called without such args, 
> can be just as convenient as if there
> were no such args possible.

Yes indeed; and I agree with what you wrote in your longer second mail
about that. They do lessen the need for more functions which means less
API surface to document and potentially lookup and learn. Isn't a good
example that window functions that I wanted to make into same once in
the context of help/info patch? I don't remember which it were tbh
:). Something with selecting window on just current frame or all frames
or something like that.