James Devenish wrote: > But, this still leaves the question...how come the machine dies when > &status is invalid? A guess: 'cos Linux threading is sucky, and in particular the new threading system (NPTL) used in 2.6 is new and probably less tested than it could be. Cameron.