Markus Schöpflin wrote:

Jonathan Turkanis wrote:

...

The iostreams tests which depend on Boost.Random are now failing, despite the fact that both of the Boost.Random tests pass. See, e.g., http://tinyurl.com/bozm9. I gather this is because of the recent changes to the toolset.

Any ideas how to fix this?

See the recent threat titled "[config] Why does BOOST_NO_INTEGRAL_INT64_T depend on BOOST_HAS_LONG_LONG?"

Thanks, I wan't following that thread. Jonathan