Non-reproducible results when using same seed but running on different machines

[fbradascio]

Error encountered for the calculation of the post-trial p-value. Simulations seeded with the same random number (given as input) give different results when running locally or on different machines on the DESY cluster.

[robertdstein]

This is probably not exactly a bug, but more likely a fundamental consequence of different precisions on different machines. More study is needed though to confirm.