Re: [boost] [review] [Outcome] Deniz' review

Subject: Re: [boost] [review] [Outcome] Deniz' review
From: Niall Douglas (s_sourceforge_at_[hidden])
Date: 2017-05-24 19:50:11

> (Maybe, it is even possible instead to provide some hints to
> compiler-vendors where to try to improve their optimizers?)

Back some years ago I submitted bug reports with repros to clang and
MSVC regarding the low quality of optimisation with Outcome. I have no
idea if the repros were used, but the most recent versions of both
compilers do significantly better than previous versions.

>> - Do you think the library should be accepted as a Boost library?
>>
> Yes, I would say so.
> This would be an unconditional yes, even though I would suggest to add a
> small "introductory-tutorial" to the documentation and would like to
> hear what Niall (and others) think about my above remarks concerning
> removal of "option" and re-checking of the specific compiler-optimizations.

Option is just a typedef, very easy to remove.

Andrzej I think just volunteered to redo the landing for me to match
what you want.

Thank you for your review!

Niall

-- 
ned Productions Limited Consulting
http://www.nedproductions.biz/ http://ie.linkedin.com/in/nialldouglas/