From: DEAN STURTEVANT (dean_at_[hidden])
Date: 2001-01-05 13:57:51
I think the presence/absence of hash_set etc. is better handled by a
define such as BOOST_HAS_HASH. This is more principled in the sense
that hash is not in the C++-standard, and can more easily handle
cases where a non-compiler-vendor-supplied STL is used (e.g., the
Right now, boost\config.hpp doesn't really handle this case that well.
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk