New submission from Yury Selivanov <yseliva...@gmail.com>:
I've had a few conversations with people who were confused that asyncio starts to behave weirdly when a ProcessPoolExecutor is set as the default one. We don't really test that asyncio's built-in functionality (like DNS resolving) works well with a process-pool, which leads to bug reports like [1]. Third-party libraries also always assume that the loop is always configured to use the ThreadPoolExecutor (as it is by default), and also don't even test against ProcessPool. My idea here would be to deprecate setting ProcessPoolExecutor as a default one in 3.8 and prohibit that in 3.9. Guido, Andrew, what do you think? [1] https://bugs.python.org/issue34073 ---------- components: asyncio messages: 321324 nosy: asvetlov, gvanrossum, yselivanov priority: normal severity: normal status: open title: We should prohibit setting a ProcessPoolExecutor in with set_default_executor type: behavior versions: Python 3.8 _______________________________________ Python tracker <rep...@bugs.python.org> <https://bugs.python.org/issue34075> _______________________________________ _______________________________________________ Python-bugs-list mailing list Unsubscribe: https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
