Ben Darnell

#3272 is now merged, which detects the closure of the event loop with an async generator and cleans up the thread. This means that we clean things up reliably when...

@neilsh It's possible, but I don't see a problem in that area. The paragraphs you quoted talk about the importance of re-raising CancelledError if you catch it. The change to...

Yes, the design challenge here is that `finish()` itself is the main point of coordination with other systems. `on_finish()` is really a *post*-finish hook, so if you make it asynchronous,...

The typing is accurate; the docs are wrong. `WaitIterator` does not currently work with all awaitables, only Futures. (`Event.wait` returns a `Future` even though its typing says `Awaitable`) However, I...

I've found an older issue #1239 which is essentially the same as this one. I'm going to close this one as a duplicate of the older issue (although the conversation...

Actually I'm going to reopen this but retitle it to be about the behavior raised in https://github.com/tornadoweb/tornado/issues/2398#issuecomment-1071789709: This issue is now about how `python -m tornado.autoreload` handles programs that exit...

With the introduction of SameSite cookies, Tornado's `xsrf_cookies` system is obsolete. I intend to deprecate it which means that issues like this one will go unfixed (I do not intend...

I've thought about including this in Tornado 6.3, although I've tried a few designs and haven't found anything I'm happy with yet. An `enforced` boolean on `set_cookie` that adds the...

I'm still interested in having ASGI support but have very little time for Tornado development these days. FWIW, I don't think ASGI has anything to do with the performance differences...

The root cause of this issue is that `WebSocketHandler.close` is not immediate - it begins the process of shutting down the websocket connection, but this takes a little time (the...