Steam Discoverer

A tool to discover Steam clients on your network and find out what they are playing.

Steam API - Breaking Changes

Steam API now restricts who can view your game activity. The default setting is "Friends only". This means that the only users whose state you can see are friends of the account whose API key you are using, or users who have changed the profile setting away from default. This severely limits the utility of this application.

Features

• Discovers steam clients on your LAN using the In Home Streaming discovery protocol
• Looks up Steam API to see what they're playing
• Optionally stores data in DB (postgres) for later analysis
• Optionally pipes data to a redis pub/sub queue
• Live graph of what people are playing (requires redis pub/sub queue)

How it works

Steam In-Home Steaming uses a UDP discovery mechanism to find other steam clients. We use this for the first step to find out which Steam IDs are present on the LAN.

Then, we query the Steam API to find out what each Steam ID is doing. The API returns the user's display name and the current game. Hey presto, we have our data!

There are three caveats here -

1. If a user does not have steam in-home streaming enabled, we will not know that they exist
2. If a user has their Steam profile set to private, we will be able to see what their display name is, but not if they're online or which game they're playing
3. The Steam API is limited to "100k requests per hour" (or something like that, per the terms of use) - so if you're at a giant LAN party, don't set this to poll the API every 2 seconds.

FAQ

Woah, this is scary. You can, like, totally stalk me?

Correct. You can opt out of this by disabling in-home streaming. If you are the person running this utility, please do not abuse the data. Don't publish it online. Strip all identifiable data from your dataset ASAP. Be a Good Person, okay?

In fact, if you just want to get the cool graph of how many people are playing which games, you should just disable postgres (see config.json) so that no data is ever stored :)

Can I see which non-steam games people are playing in Steam?

According to the Steam API documentation, yes. But in reality, the API does not return this information, so I cannot give it to you. I'm sorry :(

Can I get this data for Origin, LoL, Battle.Net, etc?

No. Not with this tool, as this is a very Steam-specific tool. The other services have APIs, and maybe there is a way to identify users on your LAN - I don't know, I haven't looked :)

We run VLANs at our LAN. Will this still work?

Yes. You have two options.

1. Set up your router to rebroadcast UDP port 27036 packets with a source IP coming from your host running this
2. Trunk all of your VLANs to the host running this, and set up an interface for each one. This tool will broadcast out on every IPv4 interface on the system. (configuration for that coming soon)

I want to run this at a giant event with over a thousand people. What should I do? Requests to the Steam API are batched up in lots of 100. Making lots of simultaneous requests makes us bad people, so we have a limit of 2 concurrent requests to the API (you can update this in steamapipwrapper.js). Unless your internet is terrible or you are scanning 12k people, this probably won't be a concern for you.

Additionally, the SteamAPI has a limit of 100k requests per day. The default broadcast interval is 30s. 24 * 60 * 2 * (1000/100) = 28.8k. So that's safe and under the API limit. But if you ran at 10 second broadcast intervals, or your event increased to 3000, you would exceed the Steam API limits. Be careful!

Do you have any screenshots?

Not yet. It's just a pie chart and a text console. Very exciting stuff.

Roadmap

• Make library to abstract this listening - expose just "we saw a client, here's details!"
• Add more configuration options, eg, immediately anonymise data (don't record steam ID)
• Add feature to auto-anonymise data, so that privacy isn't much of a concern

Requirements

• A recent version of node. Tested with v5.7.1.
• For data storage in postgres, a postgres server (optional)
• For the live visualisation service, Redis (optional)

Installation

npm install

cp config.example.json config.json

# Make sure to insert your own Steam API key and postgres connection info in here!
# Get your Steam API key from http://steamcommunity.com/dev/apikey
editor config.json

# Connect to your postgres DB, create a user and database, and run the postgres-init.sql script
# (or mysql-init.sql script if you're using MySQL)

node app.js

In this configuration, data is saved to a postgres database for later analysis.

Using the config.json file, you can disable postgres if you don't want to save the data. You can also enable Redis to enable the usage of the live updating graph (see next section)

Data Sinks

In the configuration file, you can specify a number of places to send data (sinks). These are currently:

• Postgres
• MySQL 5.7.8 and up (requires JSON data type)
• Redis
• UDP

You can specify each one multiple times if, eg, you want to send to multiple UDP ports for whatever reason.

Using Postgres, you will need to initialise a database with the postgres-init.sql script.

The Redis sink is currently used for the visualiser (see below), so only sends a very limited subset of data to the pub/sub queue. If someone asks, I'll make it send all data to configurable pub/sub queues too.

The UDP sink can be used to send the data to your own application or log processing applications like Logstash.

I plan to add TCP and Syslog in the near future.

Excluding fields

If you wish to exclude a certain field entirely or from a specific data sink, you can do so in the excludeFields configuration option. This can be done globally or per-sink. See the example config file. Names are case sensitive.

Live updating graph

The visualiser directory contains another small app that provides a web page with a live updating graph of which games people are playing.