Re: Boost vs Std library features (was: Re: [scope] Scope review starts on November 26th)

From: Jeff Garland (azswdude_at_[hidden])
Date: 2023-11-29 02:56:42

On Tue, Nov 28, 2023 at 7:24 PM René Ferdinand Rivera Morell via Boost <
boost_at_[hidden]> wrote:

> On Tue, Nov 28, 2023 at 7:34 PM Peter Dimov via Boost
> <boost_at_[hidden]> wrote:
>
> > LEWG is not an efficient mechanism for designing libraries, because
> > iterating here is vastly faster and better than iterating at WG21
> meetings.
>
> I agree with the conclusion that WG21 should stop designing libraries.
>

It's clear LEWG isn't designing libraries.  They're completely dependent on
authors to bring a design and modify it in a way that makes the committee
happy.  My conclusion is that LEWG isn't great at reviewing libraries --
and with libraries going around Boost they miss the review and likely the
user base that confirms that the library makes sense.  Note that in my
experience a decent percentage of the library work is actually modifying
the existing library.

> And should also stop contemplating non-foundational libraries.

Please define what you think that is. Would std::process qualify, because
I'd like to think it does? I'm assuming you'd agree that
std::format/std::print are important?

> But
> until there's a reasonable library ecosystem using STD as a
> distribution mechanism will continue. Boost happens to be one of the
> better such ecosystem distribution mechanisms, still.
>

It's fine -- but I'd argue that there's a lot of people that avoid boost
like the plague.  It doesn't matter if that's fully rationale, but it's a
perception that clearly is out there.

Jeff