Re: [boost] [concept_check] Standalone boost::ignore_unused_variable_warning()

Subject: Re: [boost] [concept_check] Standalone boost::ignore_unused_variable_warning()
From: Adam Wulkiewicz (adam.wulkiewicz_at_[hidden])
Date: 2014-05-31 08:18:41

Hi,

2014-05-20 12:45 GMT+02:00 Adam Wulkiewicz <adam.wulkiewicz_at_[hidden]>:

> Hi,
>
>
> Rob Stewart wrote:
>
>> On May 19, 2014 11:26:49 AM EDT, Adam Wulkiewicz <
>> adam.wulkiewicz_at_[hidden]> wrote:
>>
>>> Then probably also
>>>
>>> boost::ignore_unused_variable(v1) ;
>>> boost::ignore_unused_variables(v1, v2, v3);
>>>
>>> for consistency.
>>>
>> Actually, I'd favor ignore_unused(). I see no reason for "variable" to be
>> in the name. If you use a macro, then you'd need to distinguish between
>> TYPEDEFs and VARIABLEs.
>>
>
> I like those new names but don't forget that function
> ignore_unused_variable_warning() is already used around Boost. Grep shows
> ~600 uses in the code of various libraries and ~100 in the docs.
>
> And to be honest, I like this long name. When I read it I have an
> impression that it's not a part of the algorithm. But maybe I got used to
> it.
>

Ok, lets finalize this!

So, do you think we should:
1. Pick the new names and leave boost::ignore_unused_variable_warning() in
Boost.Concept
2. Pick the new names and move the old implementation from Boost.Concept to
the same place?
3. Use the old name and remove the old implementation from Boost.Concept?

Btw, Rob I think your proposal is very elegant:

boost::ignore_unused(v1, v2, v3);
boost::ignore_unused<T1, T2, T3>();

>
> I'm also curious what the author/maintainer of ConceptCheck thinks about
> it? Jeremy are you reading this thread?
>

I guess not.

Regards,
Adam