Thomas Eizinger

Ah interesting! That makes sense. Thanks for sharing. I wonder if we can make the Internet site special here and introduce regions just for that one before adding it to...

This sounds like we need to patch `if-watch`. From an interface perspective, a stream ends when it returns `None`. It is not wrong per se to have a stream return...

> Could this be in any way related to an issue I'm seeing on Android where interfaces seem to be dropped/readded automatically on a loop? Basically a SwarmEvent::ExpiredListenAddr followed by...

> I haven't done any deep dive into the if-watch code to understand the issue fully and patch easily but I can certainly try. A couple pointers would be appreciated,...

> @thomaseizinger turns out the issue I'm seeing on android is unrelated ([mxinden/if-watch#30](https://github.com/mxinden/if-watch/issues/30)), but I would have taken a way longer time finding it if you hadn't pointed me in...

@mxinden Do you have any input on this?

> > In reality, the set of transports we can perform address translation on is limited by the transports that the `multiaddr::Protocol` enum supports. Thus, delegating this to the actual...

> > If we have `port_reuse` enabled in the transport, these ports will already be the same > > I am not sure this is true. In the case of...

So if I understand correctly, the only blocker for this is that `Swarm` needs to somehow (reliably) learn that `libp2p-tcp` was configured with port-reuse. I'll shelf this for now but...

Technically, the setting for port-reuse is per connection right? In theory, a transport could allow for changing this at runtime. Thus, a better model might be if `Transport::Dial` where to...