Closed oskarth closed 3 years ago
cc @kdeme @decanus
Brief update: basic scripts work.
Some blockers: https://github.com/status-im/nim-waku/issues/147 https://github.com/status-im/nim-waku/issues/148
@decanus @kdeme I'd appreciate you helping out with getting this testnet out and working.
Issues before addressed, main blocker is this: https://github.com/status-im/nim-waku/issues/155
Spec issue slightly separate but there's an issue here: https://github.com/vacp2p/specs/issues/178 I'll look at it tomorrow.
Above blocker resolved.
Blocked by @decanus and @kdeme also running it so we can test end to end.
Spec draft PR here https://github.com/vacp2p/specs/issues/178 but once above is addressed we can close this I believe.
Tested and works.
Specs merged too, closing as done.
This is the first internal testnet for Waku v2.
Goals
Scope
Scope is as minimal as we can get away with. This is deliberate, in order to get it running as soon as possible.
Not in scope:
Where does the name come from?
Nangang (南港) Station is a metro station in Taipei. https://en.wikipedia.org/wiki/Nangang_station
When will it start?
Plan is for this week, probably Wednesday or Thursday. This assumes no major blockers are found.
How long will it run for?
As soon as a quantum of utility exist for other features/infrastructure, so probably just a week or so.
Tasks
1) Ensure two nodes can connect to each other directly and use publish/subscribe to talk -- connecting works with quicksim so should be fine -- e.g. with examples/quicksim style RPC 2) Ensure two nodes can discover each other via third node 3) Deploy node to public place (cluster or ad hoc) --- make sure this is known location that can be hardcoded (bootstrap/static node) 4) Sanity check for developers to talk over a single topic 6) Update spec to draft mode 7) Mini update people about this (Vac forum/tweet) -- no big thing, this is embarrassingly limited in scope (on purpose)
...fill in more in comments/edit issue directly.
Acceptance criteria
A testnet of Waku v2 is running that:
WakuMessage
s overrelay
protocol for a specific topic -- i.e. publish and subscribe with basic RPC -- incoming messages can simply be logged to STDOUTIf we can get basic
store
andfilter
support that'd be great, but it isn't going to block the launch.