Re: [boost] [xint] Boost.XInt formal review

Subject: Re: [boost] [xint] Boost.XInt formal review
From: Artyom (artyomtnk_at_[hidden])
Date: 2011-03-08 00:55:26

----- Original Message ----
> From: Scott McMurray <me22.ca+boost_at_[hidden]>
> To: boost_at_[hidden]
> Sent: Tue, March 8, 2011 3:53:38 AM
> Subject: Re: [boost] [xint] Boost.XInt formal review
> 
> On Mon, Mar 7, 2011 at 13:06, Artyom <artyomtnk_at_[hidden]>  wrote:
> >
> > - Move Semantics:
> >
> >         boost::xint::integer a,b;
> >        a = 100;
> >        b =  boost::move(a);
> >        std::cout << a<<" " <<b  << std::endl;
> >        // expect 0 100, get 100, 100
> >
> >   integer is container and should behave like container
> >
> 
> I think  that fact should be hidden as much as possible.
> 
> What do you get from int  a = 100; int b = boost::move(a);?
> 
> I'd assume 100, 100, so that result  from xint is a feature, not a bug.
>  (Though obviously relying on that fact  would be a terrible idea.)
> 

The point that same happens even for big integers, so I expect that move
would prevent needless memory allocation.

Finally the performance matters, so such optimizations are important.

Artyom