I'm making an application where I need to map sockets to some object. I thought I would be safe using the socket.id variable provided by Socket.IO. Unfortunately while debugging I found out that the socket.id changed without the client disconnecting/reconnecting or whatever.
Here's a small piece of code to show what I'm trying to do:
var io = require('socket.io').listen(8080),
var myVariable = new Array();
// Main event loop
io.sockets.on('connection', function (socket) {
socket.on('myEvent', function(data) {
myVariable[socket.id] = data;
}
// 'someOtherEvent' which uses the mapped data.
socket.on('someOtherEvent', function(data) {
// Doesn't always work because socket.id has changed and var is empty
doSomethingWith(myVariable[socket.id]);
}
});
I'm not sure but I don't think this is the desired effect. Why does this happen and how would I work around it?
I was using node-inspector to debug my application and because I was holding the code on some breakpoint it disconnected the client (probably some timeout client side). That's why I got a new socket.id when I continued the code execution.
Looks like 'someOtherEvent' emited before 'myEvent' or socket has been reconnected. I use socket.id as connection identifier on two different projects in production and it works without any problems.