> one more idea, a kind of named loop: interesting idea, thanks.
> > When it become too complicate, I use state > machine:http://en.wikipedia.org/wiki/Finite-state_machine I unsuccessfully played a bit with a FSM, but there is a lot of data that is passed around between the states and a lot of counting (like trying a certain step n times), so the FSM turned out to be even more complex. And I have to keep the code simple for non CS people to run the actual experiment. The loops are kind of self-explanatory, this is exactly how you would specify the experiment, even though I am really hitting a wall at the moment. Maybe I am really missing an obvious solution, because breaking out of nested loops really doesn't seem like anything fancy. Fortran/c/c++/ Ruby/Perl all have that facility, even Java has named loops. -- http://mail.python.org/mailman/listinfo/python-list