Re: [boost] Bugs in boost random number distributions

From: Beman Dawes (bdawes_at_[hidden])
Date: 2006-08-15 06:53:30

On 8/14/06, Jens Maurer <Jens.Maurer_at_[hidden]> wrote:

>  >> (3) uniform_int<long long>(low, high) produces a distribution with the
>  >> wrong mean. E.g.,
>
>
> "long long" is not (yet) C++, thus I don't consider fixing this
> high priority.

Every modern C++ compiler supports long long (or better yet,
int_least64_t), and long long is non-controversial and has been voted
into the WP for the next standard. So people use long long and expect
to get reasonable results. Thus I'd be inclined to treat any long long
bug just as seriously as a plain long bug.

--Beman