4 ms·
I've always found Armin's posts to be intellectually honest and thought provoking. And unlike your ad-hominem comment, he itemizes his concerns and provides de
by raymondh 10y ago
I've always found Armin's posts to be intellectually honest and thought provoking. And unlike your ad-hominem comment, he itemizes his concerns and provides details (such as the performance comparison with David Beazley's curio project).
At the very least, it is a warning sign that a notable and highly experienced Python expert is having a hard time grappling with the best practices (or even workable practices) for a significant new feature set: "I know at least that I don't understand asyncio enough to feel confident about giving people advice about how to structure code for it."
As far as I can tell, not a single respondent to this thread has indicated that to the contrary, they have been able to say that they do feel confident enough to give people advice on how to structure code with asyncio.
At the very least, that means that we have a documentation and communication problem which is either intrinsic to the new API or something that will work itself out over the next few years.
- ubernostrum 10y agoFor what it's worth, I read Armin's critiques, but I take them with a grain of salt; it's clear that what he wants from a language, and what other people want from Python, diverged a while back and are probably irreconcilable at this point. That doesn't mean he doesn't have good points, but does mean that I read his articles through a lens of "the language he really wants probably is never going to be Python again".
- the_mitsuhiko 10y ago> I read his articles through a lens of "the language he really wants probably is never going to be Python again" I always had that opinion that what I would like Python to be like is never going to happen. This is not something new with Python 3.