Re: [boost] random number library review

From: Csaba Szepesvari (szepes_at_[hidden])
Date: 2000-06-08 09:23:49

> At 11:35 PM 6/7/00 +0200, Jens Maurer wrote:
>
> >Csaba Szepesvari wrote:
> >> 3) checking conformance
> >>
> >> When an rng is passed along, the client might want to have a means
> for
> >> checking if it received an appropriate one (appropriate
> distribution).
> >
> >Ok.  The plan is to have an enum
> >enum random_distribution { random_uniform_int, [...] random_normal }
> >and an appropriate "static const" member within each distribution.
> >
> >Or do we want a traits class (more extensible than a member) like
> >this (similar to std::iterator_traits<>)?
> >
> >template<class T>
> >class random_traits {
> >  static const random_distribution = ...
> >};
>
> My general sense is that the random number library is already quite
> feature rich, particularly from the standpoint of a casual user.  If
> you try to solve all problems you will end up with something only
> experts will use.  A problem presented as "the client might want..."
> doesn't sound serious enough to warrant more than a very simple
> change, if any change at all.

I think, anyone (like me) who develops a general piece of software that uses
random random generators needs this feature. Not having this feature means
that you cannot safely use these random number generators in a library that
gets them as template arguments, breaking the encapsulation view.

By the way, re the parameter passing/copying issue if you forbid the copy
constructor then how should I write a function that uses a random number
generator (with a specific distribution)? E.g. imagine I want to write a
library supporting Monte-Carlo integration: I need to check the distribution
(first issue), but I would like to use a user supplied rng because of
independence and because users might have preferences for certain rngs.

Cheers,

  Csaba