What does everyone think about updating the core client/server packages to QUIC?
Have there been any internal discussions about this? Or any reason that this would be a non-starter?
How’s the browser support of quic?
Starting to become more adopted. We could implement a support check on the client side prior to initializing a connection with the server.
I’m sure that this would be useful to some people, but I know of a readily available way of getting the benefits of QUIC today.
All of my Meteor webapp deployments use nginx as a reverse proxy to terminate the HTTPS endpoint and serve static assets as it handles these tasks more efficiently than Node.js. nginx is configured to pass requests onto Meteor using UNIX sockets.
I would be able to get the performance benefits of QUIC today by replacing nginx with nginx-quic which is currently in beta. I will definitely be enabling QUIC once nginx-quic has reached release quality and is merged back into nginx.
2 Likes
I’m for it, but I think we first need to do this:
Agreed. I was browsing through the Meteor 3.0 roadmap earlier this morning and found that discussion.
Take the Meteor 3.0 roadmap with a big reserve as right now it is just a wish list.
Okay – I am working on cleaning up / updating the core DDP packages to familiarize myself with the connection logic.
3 Likes