Benjamin Saunders

What leads you to believe that this behavior is incorrect? My guess would be that this is exactly the draining period of the attempted connection. Waiting for that to complete...

You don't have to; `wait_idle` is strictly optional. It does serve a useful purpose: it helps the peer dispose of any state it may have already allocated in a timely...

Why would we be able to use a shorter timeout? As far as I can go, a client connection with a handshake in progress isn't in any way special as...

Thanks for the report! I concur that this shouldn't have been affected. Most notably, it's a `quinn-proto` test, and those don't involve tokio at all. You could probably avoid exposure...

> Is there interest in seeing TX support via sendmsg_x? We found there wasn't much performance benefit, and was considerable difficulty taking advantage of, `sendmmsg`-style batching. IIRC the `_x` functions...

Those tests tend to be extremely noisy, as the huge variance suggests. A targeted quinn-udp benchmark might be more useful.

Probably because cargo installs the newest compatible version of a dependency the first time it's required. Run `cargo update`.

Yep, that's why the issue's still open. If it's important to you, you could make a PR.

Can you elaborate on "mem usage from crypto"? I don't expect cryptographic operations to consume significant memory, or really anything but O(input). Usually the concern with handshakes is CPU usage,...

> We're looking specifically at x509 verification, as it happens before we get the future returned to us when we accept() making it a prime target for rate limiting given...