From: Bruno Lalande (bruno.lalande_at_[hidden])
Date: 2008-05-22 06:11:16
> At the cost of one additional template instantiation (power<0> will
> always be instantiated) we get rid of the second parameter and have no
> runtime evaluation, as advocated in the other thread.
Oops... I realize that this falls back to the solution initially
proposed by John Moeller and that had been rejected by Hervé due to
the multiplication by 1. Sorry :-)
Bruno