From: Mat Marcus (mat-lists_at_[hidden])
Date: 2008-04-05 21:10:32
On Sat, Apr 5, 2008 at 5:53 PM, Peter Dimov <pdimov_at_[hidden]> wrote:
> Mat Marcus:
> > This strengthens my view that _SECURE_SCL for release builds as a
> > misfeature, and I still am of the opinion that, at the least, boost
> > build should disable 'secure' STL for release variations in the
> > default case.
> If we do that, we'll probably have to encode the _SECURE_SCL setting in the
> .lib names, to avoid mismatches.
Yes, agreed. I was envisioning a link-incompatible boost-build feature, e.g.
which defaulted to off for release builds.
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk