Re: [boost] Re: Submission: typelist

From: David A. Greene (greened_at_[hidden])
Date: 2001-11-29 17:43:16

Eric Friedman wrote:

> *_wrapper seems like a step in the right direction for the reasons
> Matthew articulated, but a "wrapper" connotes (to me at least) some kind
> of functionality being added by the wrapper to the existing type --
> which is certainly not what's happening.

Good point.

> So what about *_tag (type_tag, int_tag, bool_tag, ...)?
> 
> It's shorter, yes, but it also seems to better connote what it used for.

I kind of like this suggestion.  It gets at the fact that
type2type<type> != type and it suggests a mapping operation is
involved, with *_tag being the "index" or "iterator" where
"dereferencing" *_tag (via ::type, ::value, etc.) returns
the thing mapped.  It also suggests a lightweight "thing"
that represents a heavier object (i.e. when used for static
dispatch).

"Tag" also implies the uniqueness which I was trying to
convey with unique_type<N>.

It unfortunately clashes with iterator_tag, but maybe those
aren't the best-named things either.

                         -Dave

-- 
"Some little people have music in them, but Fats, he was all music,
  and you know how big he was."  --  James P. Johnson