Géry <gery.o...@gmail.com> added the comment: @Andrew Svetlov
> Adding a new state for "not running and not pending but something in between" > is useless I have not suggested that. I have just reported that when the number of submitted calls is strictly greater than the number of pool worker processes, the number of RUNNING calls returned by the method `Future.running()` makes no sense. Probably because the current `ProcessPoolExecutor` implementation uses 2 PENDING stores (the `_pending_work_items` `dict` and the `call_queue` `multiprocessing.Queue`) but treats the second store as a RUNNING store, contrary to the `ThreadPoolExecutor` implementation which has only 1 PENDING store (the `_work_queue` `queue.SimpleQueue`). The proper thing to do would be to correct the current implementation, not to create a new future state of course. ---------- _______________________________________ Python tracker <rep...@bugs.python.org> <https://bugs.python.org/issue37276> _______________________________________ _______________________________________________ Python-bugs-list mailing list Unsubscribe: https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
