From: David Abrahams (dave_at_[hidden])
Date: 2002-09-30 10:35:23
I guess I wasn't thinking clearly. Please don't move the directory until
I've investigated further.
I now don't see how we were getting a conflict unless $BOOST_ROOT/boost was
somehow getting into the #include <...> search path.
The directory in question is $BOOST_ROOT/boost/numeric, and only
$BOOST_ROOT should be in the #include<...> path.
I'm going to figure out what's happening.
David Abrahams * Boost Consulting
dave_at_[hidden] * http://www.boost-consulting.com
----- Original Message -----
From: "Eric Woodruff" <Eric.Woodruff_at_[hidden]>
Sent: Monday, September 30, 2002 11:47 AM
Subject: [boost] Re: Re: Possible emergency: boost/numeric
> In any case, the user must adopt a standard of usage, such that they do
> allow both <numeric.hpp> and <boost/numeric.hpp>. Documentation can
> certainly influence their decision in adopting that standard. (Not to
> mention the fact that by default, boost is installed in places like
> /usr/include/boost making <boost/numeric.hpp> the preferred and
> convenient/automatic usage. In this case, it only hurts you to deviate
> that standard.)
> I can however see an argument for a boost/numeric library that is meant
> replace <numeric> by providing some bugfix or better/more
> advanced/convenient usage as a direct extension of the standard. If such
> library deserves name reservation, then the current boost/numeric
> that reservation.
> "Neal D. Becker" <nbecker_at_[hidden]> wrote in message
> > It *is* a problem to have an identical name, because someone could use
> > -I blah and get a very unexpected result.
> > Having a name identical except for some suffix is not a problem for a
> > machine, but still may confuse humans.
> > _______________________________________________
> > Unsubscribe & other changes:
> Unsubscribe & other changes:
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk