vyzo

The seen cache would be most severely impacted, as messages can be rebroadcast into the network way after the 120s cache duration. We need to consider the effects of this.

In terms of structure, we can add an api for forwarding prepared messages (ie messages published by someone else, already signed). This way we can offload the message for async...

#176 supports long-running validators in the simplest possible manner: It removes the default (short) timeout and allows the validators to run arbitrarily long without any need for api changes or...

Note that you need to adjust the time cache duration accordingly. On the other hand there is still a use case for completely offline validators, which could take days to...

Yeah, it would be nice to factor out of band connections. Either approach works, polling probably easier to work with.

It is a little more complicated than that, but we are aware that there is a need to implement routers outside the library. We will work to address the issue...

#199 performs the necessary refactoring; it should be sufficient to implement out of tree routers.

These arguments are quite convincing, and i agree pretty much. Still, this is something steven might feel very strongly about, so lets wait on him.

> I'm not really sure if we need [dfd81e8](https://github.com/libp2p/go-libp2p-pubsub/commit/dfd81e80848e5ce4e64fbf30cfdaafb4dc56b9d3) or not. I guess adding a non-standard flag to `pb/trace.proto` would break some users of the tracers? Yeah, let's not break...

yes, looks reasonable so far.