GitLab

and silence some warnings

Issue 199

Fixes issue #200 

See merge request !19

- Run same simulation with the same configuration (seed=const)
- Change for every iteration the number of threads
- Plot evacuation times vs number of threads
- Test fails if std(evacuation times) > 0

- We need to use barrier after force/velocity calculations and
  just after the barrier start the parallel update

- The end and start values dependent on the thread ids, lead to
  race conditions if num of pedestrians is smaller then number of
  threads (pedestrians get deleted with forthcoming simulation)

- @todo: other models need the barrier mechanism too. not done in this
  commit

- @todo: Dispense with manually distributing the threads and use
  a "#pragma omp for"

Todo for other models as well

Add inputfiles/Sources

Changed router, since global router has visibility-problems

0.5 can be used now. (jpscore now works again with old compilation of
jpsvis)