Centrifugo VS Mercure

Hello, I am author of https://github.com/centrifugal/centrifugo. Our users can choose from WebSocket, EventSource, WebTransport (experimental stabilize in the future). WebRTC is out of scope as the main purpose is central server based real-time json/binary messaging, and WebRTC makes things much more complex since it shines for peer-to-peer and rich media communications.

What I'd like to add is that Centrifugo also supports HTTP-streaming – not mentioned by the OP – but this is a transport which has advantages over Eventsource - like possibility to send POST body on initial request from web browser (with SSE you can not), it supports binary, and with Readable Streams browser API it's widely supported by modern browsers.

Another thing I'd like to mention about Centrifugo - it supports bidirectional WebSocket fallbacks with EventSource and HTTP-streaming, and does this without sticky sessions requirement. I guess nobody else have this at this point. See https://centrifugal.dev/blog/2022/07/19/centrifugo-v4-releas.... Which solves one more practical concern. Sticky sessions is an optimization in Centrifugo case, not a requirement.

If you are interested in topic, we also have a post about WebSocket scalability - https://centrifugal.dev/blog/2020/11/12/scaling-websocket - it covers some design decisions made in Centrifugo.

Check out https://github.com/centrifugal/centrifugo - it was initially designed to be a standalone language-agnostic real-time messaging server. So it may be used with Django without radical change in the existing application and using ASGI. It can also provide a much better performance if you care about it.

https://github.com/centrifugal/centrifugo

It's a complete solution, including server, admin panel and client library.

We're an European company and use OVH, Hetzner and others.

Hello! Theoretically you can take a look at https://github.com/centrifugal/centrifugo - which is a standalone self-hosted real-time messaging server. It does not have native support for Laravel and not compatible with Pusher protocol, though integrating with any backend system, including Laravel: see the blog post https://centrifugal.dev/blog/2021/12/14/laravel-multi-room-chat-tutorial, also has some helper packages:

Hello, it's also possible to design an app in a way that its core will be built with Python, but WebSocket part delegated to something external and efficient like https://github.com/centrifugal/centrifugo – the benefit of the approach is that application business logic is completely decoupled from the real-time transport layer. This may lead to a scalable design with graceful degradation. I think this is especially useful when you already have backend built with Django and need to handle millions of concurrent connections.

Centrifugo https://centrifugal.dev/ https://github.com/centrifugal/centrifugo

Are you aware of things like websockets and mercure.rocks?

It could be implemented with STOMP, or Mercure (goes well with API-Platform, written in PHP/Symfony), you could write your own with the help of nchan and scale it via Redis. If it's a web service, the best practices for operating and scaling are well established, Godot then just becomes another client.

You can try https://mercure.rocks/.

It's worth looking at something like Mercure , which is used by API Platform

Pusher might be one of these systems, Mercure is another one (which you can host yourself). Mercure has good documentation and some examples in various languages, including PHP: https://mercure.rocks/docs/ecosystem/awesome#examples.

You can use Mercure https://mercure.rocks/ , Mercure uses http2.

I actually was thinking about this when I saw https://github.com/dunglas/mercure project to become a Caddy plugin. But I did not find enough reasoning to try this with Centrifugo at that point, and still... Seems awesome from one side - tight integration with a web-server, no extra network between LB and Centrifugo. But will this be useful in practice and find its users? 🤔 That's the question I don't have an answer yet.

Polling is kind of ok (data might change between poll events), but real-time updates are better. So If you don't mind adding an extra service, then take a look at mercure.

Two more questions: - Am I to understand The Guild now recommends GraphQL Yoga over Helix? - Could something like Mercure be included in the recipes ? (this would be a nice solution for serverless)