Re: [Asset Stewardship] About the Beman Project

From: Andrzej Krzemienski (akrzemi1_at_[hidden])
Date: 2024-09-15 20:15:08

niedz., 15 wrz 2024 o 21:43 Joaquín M López Muñoz via Boost <
boost_at_[hidden]> napisał(a):

>
>
> Enviado do meu iPhone
>
> > No dia 15 de set. de 2024, Ã s 21:28, Peter Dimov via Boost <
> boost_at_[hidden]> escreveu:
> >
> > Zach Laine wrote:
> >> […]
> >> Boost wants the best C++ libraries that appeal to its reviewers.  WG21
> wants
> >> to make sure something is appropriate for the standard.  Those are
> similar,
> >> and even overlapping, but fundamentally different goals.
> >
> > It doesn't really matter. For a library to get into the standard it
> needs to
> > undergo design review by LEWG and be accepted by LEWG. Adding a Boost
> > review on top of that can never be a "win" for the author, because it can
> > only increase the probability of rejection.
>
> It would if field experience weighed more in
> WG21’s acceptance criteria.
>

But because it often doesn't, the Beman Project does indeed fulfil an
important role. If nothing else (and there is likely more, that I fail to
appreciate), it allows one to prove that what is proposed is actually
implementable. Recently, there was a situation in a WG21 triannual meeting
that the proposal for inplace_vector had to be withdrawn in the plenary
voting at the last moment because it was observed that what had been
approved by the subgroups was unimplementable. Having the library in the
Beman Project would have prevented that from happening.

I am not saying that this is all good. But it could be worse without the
Beman project.

Anyway, I now understand that the expectation I expressed earlier, that the
library authors after the hearing in LEWG should be directed to Boost is
misguided.

Thank you.
&rzej;