After about 40 to 50 completed experiments (only 1 participant each) the system becomes slow and runs into bugs (files are not found [archive-l1] , entries in registry do not exist). Also: numbering of experiments is buggy. Some numbers are simply skipped.