.... "I am asking for this ticket to be re-opened, since I can still reproduce the problem in the latest 4.5.3. version"

The PR: https://github.com/redis/redis-py/pull/2641

According to the latest comments there, the bug is only partially fixed.

> "If a request is canceled after the request is pushed onto the incoming queue, but before the response popped from the outgoing queue, we see our bug: the connection thus becomes corrupted and the next response that’s dequeued for an unrelated request can receive data left behind in the connection."

The OpenAI API was incredibly slow and lots of requests probably got cancelled (I certainly was doing that) for some days. I imagine someone could write a whole blog post about how that worked, it would be interesting reading.

Was this written by ChatGPT? Maybe it found the bug as well, who knows.