Closed tabcat closed 3 years ago
Thanks for reporting this. I am going to revert the commit and ask for the server to be updated.
I released 0.20.5
with the commit reverted.
@mburns can I have your help here? I was expecting the deployed webrtc-star
servers to be updated on push, which is not the case, so we will need to update them. However, from this issue I noticed that v2 clients will not be able to talk with v3 servers 😞
I did a small research and it seems the best way would be do to
have both a group of v2 servers and v3 servers in parallel,
and route the traffic based on the EIO query parameter (EIO=3 for Socket.IO v2, EIO=4 for Socket.IO v3)
What do you think? Is this feasible on your side? The v3 server was released in v0.20.4
Thank you @vasco-santos everything is working great now :+1:
Running the latest rendezvous server locally had worked for me in the mean time as well
Thanks for the alert/feedback @tabcat I will be looking on an upgrade for this with the infra team
FYI #292 aims to solve this
Unable to connect to public webrtc-star signal servers in what seems due to a webrtc-star patch upgrading
socket.io-client
.experienced after running
npm run test
on https://github.com/orbitdb/orbit-db-store/tree/0330c952ff8b54d7d9ccf5248a32dab4642a01d3will try running the rendezvous server locally.
error from ipfs
relevant ipfs config
js-libp2p src/transport-manager.js#L143-L172
here addrs is an array of 1 webrtc-star multiaddr from the config results is an array of 1 object: