Boost logo

Boost :

Subject: Re: [boost] [afio] AFIO review postponed till Monday
From: Niall Douglas (s_sourceforge_at_[hidden])
Date: 2015-07-21 13:20:03

On 21 Jul 2015 at 9:48, Robert Ramey wrote:

> >> Did the review start yesterday?
> >
> > Short answer: Sadly I believe not. I was unhappy with the tutorial in
> > the documentation.
> >
> > Longer answer: I've been investing 40 hours per week after work into
> I wouldn't stress about this. A Boost project different than a work
> project. At work we have to get it done more or less on time even if we
> have to compromise features, quality, documentation, or whatever.
> Here we have to get it perfect and complete - whatever it takes. It's
> an obsession. Schedule means nothing here. (Note this is development -
> maintainence is a totally separate issue.)
> This is one of the key things that makes boost different. Schedule,
> effort spent, time, personal relationships, .... everything takes the
> back seat to quality and completeness. We aspire to provide the
> definitive implementation of any given idea. This is why only we can do
> this work. This is what makes us special and irreplaceable.
> Good luck with this.

Thanks for the support Robert.

I wouldn't say we aim here to get a Boost library perfect and
complete though, rather I'd say we prioritise what ought to be
prioritised if one is aiming for the best possible engineering
(according to each of our personal opinions on that) tempered by the
scarcity of personal resources and time and family demands.

For example, this v1.4 API refactor I should have done last year, but
refactoring APIs is tedious and boring, and the old one works just
fine, it's just quite far from the Concurrency TS now so I put it

So getting ready for this review was really about motivating myself
to invest 350 hours of outside of work time in doing all the stuff I
kept meaning to do but couldn't motivate myself to normally do
because it's so tedious. Indeed, the lightweight monadic future
promises I was banging on about eighteen months ago, the big hill for
me to climb was actually writing some code to implement them as it
was always going to be a 250 hour minimum, which is tough to find
after your day job.

Still, it's done now, and like APIBind I think lightweight monads are
a real game changer for how one writes C++, not least the
heterogeneous future wait composure.


ned Productions Limited Consulting

Boost list run by bdawes at, gregod at, cpdaniel at, john at