Walter Bright wrote:
> 2) Nameless parameters come up an awful lot in C/C++ header files - I think they're expected.

I thought of that one, and since you want D to work in the same way as C and C++ when the syntax is the same, I figure you must keep nameless parameters working in the way they do.

That means that we need a modified typeless parameter syntax. I'm not sure which keywords and operators are available, but I think that @ isn't used. How about just sticking a @ into the syntax for typeless parameters to make it unambiguous.

auto foo(@x, @y, int) // x and y are typeless parameters, and the third paremeter is nameless.

Cheers,

Reiner

On Sat, 02 Sep 2006 13:14:31 -0700, Walter Bright wrote:

> Reiner Pope wrote:
>> Thoughts, anyone?
> 
> 1) Lots of people like nameless parameters to implicitly document that the parameter isn't used.

And lots of people like named parameters to explicitly document the meaning
of the parameters.

> 2) Nameless parameters come up an awful lot in C/C++ header files - I think they're expected.

So what? This is D and not C/C++. If one is going to use a new language then one should be prepared to learn a few new things too.

-- 
Derek Parnell
Melbourne, Australia
"Down with mediocrity!"

Derek Parnell wrote:
> On Sat, 02 Sep 2006 13:14:31 -0700, Walter Bright wrote:
> 
>> Reiner Pope wrote:
>>> Thoughts, anyone?
>> 1) Lots of people like nameless parameters to implicitly document that the parameter isn't used.
> 
> And lots of people like named parameters to explicitly document the meaning
> of the parameters.

It would be better to be able to explicitly document that the parameter isn't used. Maybe by reusing the 'null' keyword.

int somefunc(int x, uint null)
{
}

>  
>> 2) Nameless parameters come up an awful lot in C/C++ header files - I think they're expected.
> 
> So what? This is D and not C/C++. If one is going to use a new language
> then one should be prepared to learn a few new things too.

It would add a lot more pain to converting C header files to D. For example, the Win32 API project, would require over ten thousand changes.