|
Boost : |
From: Gennadiy Rozental (gennadiy.rozental_at_[hidden])
Date: 2006-08-11 13:33:21
"John Maddock" <john_at_[hidden]> wrote in message
news:005f01c6ba4e$97268fa0$875e0352_at_fuji...
> Gennadiy Rozental wrote:
>> Ok how about this:
>
> Yep that's OK as long as your special limits class usage is resticted to
> this relative error calculation: there are circumstances where it would do
> the wrong thing if max_value() doesn't return a large enough value, in
> this
> specific case it doesn't matter because one large value is as good as any
> other when talking about relative errors.
Updated in cvs .
Gennadiy
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk