Vito Caputo

Is it really necessary to do such awkward things for determining capabilities? There's already a greeting handshake, all we'd need is for the Editor's reply to include the version/capabilities instead...

Reuse is just tricky because the lifecycle of the track-nameID mapping is unclear. It's kind of similar to the PID reuse problem in *NIX. The trouble is GET_TRACK, despite sending...

Now we just need mechanism for versioning the protocol that doesn't break existing users...

I actually thought the same thing, but observed the long delay from connect() while testing my changes in WINE here. It only happened in WIN32 and not Linux. During development...

Nah, and that doesn't get set until after the connection is established anyways.

Ok, since I doubt we'll come to agreement on doing this better within Rocket, how about we change tack altogether: Simply add an alternative to `sync_tcp_connect()` for callers who want...

> I hacked something up in #101... How does this look? It's still a blocking connect so it doesn't address all of my issues. It's a step in the right...

> > It's still a blocking connect so it doesn't address all of my issues. > > But is that a real-world problem, or a theoretical one? As I wrote...

I'm down with enabling a reconnect without name resolution in the existing tcp sockets implementation, so this PR conceptually seems like an improvement. But I'd personally prefer having this API...