Marten Seemann

After discussing this with @whyrusleeping, I wrote a short design doc how we can make 0-RTT replay-safe: https://docs.google.com/document/d/1xBin9zQoutncY4rmCivYcQRMwCsxJ1r3dnFrzdaFkFE/edit?usp=sharing

> there might be downstream dependents. The only one in libp2p I can find is `go-libp2p-kad-dht/ext_test.go`. They'll have to find a way to deal with this. This code has caused...

I’d argue that it shouldn’t live within libp2p then (or at least not go-libp2p). How about moving it to a separate repo?

We can. The original plan was to deprecate go-libp2p-testing, but we can just shift packages around: all the muxer and transport tests go here, and the mocknet tests go there.

This is probably just another variant of this test failure: ``` === RUN TestIDService id_test.go:86: test peer1 has peer2 addrs correctly id_test.go:101: test peer2 has peer1 addrs correctly id_test.go:102: expected:...

``` === RUN TestIDService id_test.go:195: test peer1 has peer2 addrs correctly id_test.go:208: test peer2 has peer1 addrs correctly id_test.go:209: Error Trace: id_test.go:41 id_test.go:209 Error: elements differ extra elements in list...

We also need to dial done the verbosity of log produced from the go-libp2p-testing suite. Having thousands of log lines is not really helpful...

Managed to get a failure with a log trace: ``` === RUN TestWaitForCandidates 2022-04-22T17:30:14.004+0100 DEBUG autorelay autorelay/relay_finder.go:587 starting relay finder 2022-04-22T17:30:14.156+0100 DEBUG autorelay autorelay/relay_finder.go:182 found node {"id": "QmRUcxbiw2Un1YSoEfrRX5JDWNrDdPmqFHcDcC5fKcVCHf"} 2022-04-22T17:30:14.160+0100 DEBUG...

And another, unrelated one: ``` === RUN TestWaitForCandidates 2022-04-22T18:37:46.677+0100 DEBUG autorelay autorelay/relay_finder.go:587 starting relay finder 2022-04-22T18:37:46.968+0100 DEBUG autorelay autorelay/relay_finder.go:182 found node {"id": "QmWnykjv5AVsRsSxkMayLo4zt3H5tdW4AFNZZPz11bXMoW"} 2022-04-22T18:37:46.970+0100 DEBUG autorelay autorelay/relay_finder.go:232 node supports relay...

I'm late to the party, but this is something we might want to pick up at some point, so here's my proposal. > Note: If we can ensure that AutoNAT...