Re: [boost] [variant] regressions in GCC 2.95 --seemingly anMPL-relatedproblem