Subject: Re: [boost] [Boost-users] [review][constrained_value] Review of Constrained Value Library begins today
From: Mika Heiskanen (mika.heiskanen_at_[hidden])
Date: 2008-12-03 11:23:13
Mathias Gaunard wrote:
> I say it is up to the programmer to ensure valid values are provided.
> If the value is only known at runtime, there should be explicit checking
> in the code prior to assignment.
If the value must be checked beforehand, why bother using a constrained
type at all? Where's the benefit if all assignments are already guaranteed
to be valid by external tests? Might just as well use a test+POD combination.
--> Mika Heiskanen
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk