The current output of the test suite could be more machine readable. We should come to a good solution for that, especially wrt. measuring test coverage,
#50 (closed)
/cc Mathias Göbel, Frank Schneider, Michelle Weidling