result_handler vs. output_handler

I personally find the fields result_handler and output_handler too ambiguous. For me, output and result are one and the same thing.

Maybe we can reorganize/rename them into:

• storage_handlers (which handle storing resutls in to DBs, Pickles, CSVs, ...)
• result_handlers (which do something with the values returned by the evaluator, e.g. make some fancy graphs ore LaTeX tables)