Geneva Optimization 0.6 "Route E. Fermi"

The main changes in this release include:
 - Added option to GBasePopulation to stop optimizing after a user-defined
   number of stalls.
 - Allowed to selectively switch off mutations (i.e. value changes) for
   GParameterBase derivatives.
 - Gave GBrokerPopulation the ability to dynamically adapt the waitFactor_
   variable upon request.
 - Added a new selection scheme MUNU1PRETAIN to GBasePopulation. In this
   scheme, the best parent of the last generation is retained, unless a
   better child was found. The other parents are replaced by the best children.
   This is a hybrid between the MUPLUSNU and MUCOMMANU modes.
 - Added micro-training environment to GBasePopulation (and derivatives).
   When the optimization stalls, parents may perform structural updates. As
   these will very likely lead to a decrease in quality, a one-time selection
   policy of MUPLUSNU is used.
 - Added the option to set an arbitrary start generation. This is useful when
   intermediate results are written out, whose file name is based on the
   current generation. Files can continue to be written out with sequential
   generation numbers, even if the optimization was started anew, e.g. using
   a checkpoint file.
 - Added the option to prevent the data of unsuccessful mutation attempts to
   be returned to the server. The goal is to lessen the network traffic. On
   the down-side, the MUCOMMANU and MUNU1PRETAIN selection mode also need
   sub-optimal results and might work less efficient, when this setting is
   enabled.
 - Prevented GBrokerPopulation from becoming an individual inside of itself.
   Doesn't make sense and will likely not work.
 - Added the option to let Geneva automatically determine a suitable number
   of processing threads in GBoostThreadPopulation (set the number of threads
   to 0 using GBoostThreadPopulation::setNThreads() to use this feature).
 - Added support for building Debian/Ubuntu packages.