use and purpose of websocket multiplexing extension

Question

I am trying to understand the purpose of websocket multiplexing extension, the main purpose what the document states is to use a single same origin physical websocket connection to the server while each browser tab uses a logical connection multiplexed on top of physical connection. I see another alternative using which we can accomplish this behavior today, the approach is to launch a shared web worker which opens the websocket connection and let each browser tab send and receive message to/from this worker. i have not tried this yet and i wonder will this work at all.

Answer 1

Your question reflects a misunderstanding of the problem the WebSocket Multiplexing Extension is trying to solve.

The base WebSocket spec (RFC 6455) defines a protocol for bi-directional exchange of data over TCP/IP. A WebSocket starts as a normal HTTP request / response. In this exchange, the client and server negotiate to switch to the WebSocket protocol. After the switch, the client and server exchange data frames over the TCP/IP connection. This creates a bi-directional data stream between client and server.

A drawback of the base protocol is that it supports only a single stream of data flowing in each direction. The multiplexing extension augments the base protocol, by allowing the client and server to create multiple "channels" over the same TCP/IP connection.

So the purpose of the multiplexing extension is to allow multiple WebSocket channels to run over the same TCP/IP connection. That's all.

Having multiple browser tabs (or web workers) share a single TCP/IP connection is just an example of how multiplexed websockets might be used. In standards terminology, it's just "informative" (descriptive), not "normative" (a required part of the spec).