This is an implementation of the lockstep synchronization architecture. Lockstep is often used in games for synchronizing large world states in real-time across jittery networks like the internet.
I wrote this for fun and I have not taken the time to make it 100% production-ready. It should be pretty fast though. It easily syncs a world with 1024 units between four players almost without registering on my CPU or network monitors (Early 2013 Macbook Pro).
I would not recommend using the code in your own project. Rather, I hope a few people can learn from it by studying it.
Have a look at this demo configured to 4 players and 1024 units: GIF.
The project uses good old make. Before you can call make you must create the file osx_project/MakefileSettings with something like the following:
BUILD_DIR = ~/Build/Lockstep
Use this file to specify where you want the products to be built.
Then you can build all the targets via the command make [TARGET_NAME]. Your current working dir must be osx_project. Here's the most important targets:
- all = build all targets
- debug_server
- release_server
- debug_client
- release_client
- test
- rt = compile and run unit tests
- rdp = compile and run debug pair (pair means both client and server)
- rdc = compile and run debug client
- rds = compile and run debug server
- rds = compile and run debug server
- rrc = compile and run release client
- rrc = compile and run release client
Per default the server only expects one player. For example, you can write make PLAYER_COUNT=4 rrs to make it wait for 4 players/clients. The PLAYER_COUNT param can also be passed to the resulting server binary of course.
Since this is just a project I made for fun I have not added support for other platforms. But the code is arranged so that it should be very easy to do. For example have a look at client/osx_main.mm. This file only implements the platform "wrapper" and everything else is delegated to platform agnostic logic.
For more information about lockstep synchronization I recommend these: