You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
It was like this before in waku_node.nim and now in waku_lightpush/callbacks.nim . But given a mounted relay is needed for a lightpush service node, it is a question of why we would want to still mount it and not fail early.
Acceptance criteria
Gather consensus on whether we would want to fail early and not start a missconfigured node with lightpush and without relay (@jm-clius@SionoiS@gabrielmer )
If we agree on doing it, then go ahead with implementation
The text was updated successfully, but these errors were encountered:
No strong opinions here, but conceptually we've decoupled lightpush from having any other protocol dependencies - in other words, lightpush is now simply the protocol for directly pushing a message to a service node. The services provided by this service node is currently limited to Relay, but it could in future be any other service (such as Store, etc.). Lightpush should then be mounted without reference to other protocols and only individual lightpush requests should succeed or fail based on the secondary protocols and services requested by the client.
It is right, but see the case if there is a running service node in the network having just able to reply error.
It will be the same for such case if dependent services are not present.
My intention was with the comment made on Akhil's changes that if I configure a node to have lightpush might want to notice early that I miss-configured my node. In this sense I mean to fail rather early than having bulk error messages for possible clients.
Background
This is a follow-on for a point raised by @NagyZoltanPeter at regarding mounting lightpush without relay.
Details
It was like this before in waku_node.nim and now in waku_lightpush/callbacks.nim . But given a mounted relay is needed for a lightpush service node, it is a question of why we would want to still mount it and not fail early.
Acceptance criteria
The text was updated successfully, but these errors were encountered: