Why is WakuRelay a protocol different from GossipSub? #309
Replies: 5 comments
-
This indeed creates some issues when trying to use js-libp2p-gossipsub. For example, as per gossipsub behaviour, when a peer subscribe to a topic, it adds know peers that are already subscribed to the topic to a mesh. This does not happen because a check on whether these peers use gossipsub is done:
Which means waku peers are not added to the mesh and down the line, no messages are sent to those waku peers. I can workaround it but I just want to highlight that this choice does create some obstacles. |
Beta Was this translation helpful? Give feedback.
-
Copy-paste from chat:
|
Beta Was this translation helpful? Give feedback.
-
Additionally, the goal with semantics is also to capture these notions #87 |
Beta Was this translation helpful? Give feedback.
-
@richard-ramos has this been answered? |
Beta Was this translation helpful? Give feedback.
-
Generic message follows I'm going to go ahead and close this discussion as:
If you feel like this is in error, please create a thread on Vac forum or a new issue. |
Beta Was this translation helpful? Give feedback.
-
While attempting to implement WakuRelay in go-waku, i noticed that go's libp2p-pubsub implementation does not support extending the list of protocols (see discussion here: libp2p/go-libp2p-pubsub#408), which increased the difficulty of implementing this spec in go.
Is there a reason why is WakuRelay defined as a new protocol (
/vac/waku/relay/2.0.0-beta2
) instead of using GossipSub directly (with/meshsub/1.1.0
and1.0.0
) + StrictNoSign? From reading the spec, the intention behind WakuRelay being a new protocol is not clear to me.Beta Was this translation helpful? Give feedback.
All reactions