Re: [boost] implicit_cast inventor

From: Peter Dimov (pdimov_at_[hidden])
Date: 2002-12-02 07:01:12

From: "Beman Dawes" <bdawes_at_[hidden]>
> At 11:26 AM 11/30/2002, David Abrahams wrote:
>  >"Peter Dimov" <pdimov_at_[hidden]> writes:
>  >
>  >> Google tells me that I was wrong, and the "correct" implementation was
>  >> originally suggested by C. Green in
>  >>
>  >> http://lists.boost.org/MailArchives/boost/msg00886.php
>  >
>  >Hmm. Can anyone track him down to get permission to use it? I don't
>  >recall seeing his name in any recent postings.
>
> Mmmm... While I think it is great if you discuss it with him or her, and
> I've sent you the email address privately, I also think that simply
> crediting him or her is enough.

I agree. Asking permission is a nice gesture, of course, but nothing more.
In this particular case, while the implicit_cast implementation appears in
its entirety in C. Green's post, his/her contribution is limited to the
identity<> technique of creating a nondeduced context, a technique that I've
seen independently "invented" several times (I even "invented" it myself
once.)

>From what I read in the Boost archives, the key feature of using the
argument to perform the conversion has been credited to Valentin Bonnard.
The "implicit_cast" name and semantics have been, I think, originally
proposed by Bjarne Stroustrup. And several participants in the discussion
(Beman Dawes, Darin Adler, and others) did contribute in ways that cannot be
measured effectively by the "lines of code" metric.