[Redacted] FPS Game Title

 

Edit: 2014-07-20

While I am not permitted to talk about (particulary the art content/thematic setting of) this game due to IP restrictions, the school at which I was a student has legal permission to use this game as part of marketting/promotion. I have also (historically) seen screen shots of the game on the website of the commercial game engine used. A video made by Media Design School (together with janky debug camera/frame rate) can be found here.

This project was completed over two years ago now (well a vertical slice anyway: one single player/tutorial level and one multi-player 'capture the flag' style arena). I was responable for coding the vast bulk of the networking sub system (on top of a RakNet transport layer). 16 Player LAN works just fine, but more coding work towards the prediction algorithms (which also touches game physics) would be required to get this working well over the internet (higher latency).

For those interested in creating something similar, the GPL'ed quake source code together with notes by Fabien Sanlard are a great starting point. Using Quake 2 style networking (client player objects send state continuously via UDP, with reliable 'acknowledged UDP' messages from server for specific world events) is faster to get working, but creates a lot of race conditions that the client code has to handle. Quake 3 serializes the differences in the world (from each clients last acknowledged perspective) as an atomic block removing this complexity from the client, but this uses more memory on the server and is non trivial to set up/get working and maintain.  Ultimately with current availability/affordability of game engines like Unreal Engine and Unity those wanting to make games would be better served focusing on making games rather than engines (or engine sub systems/components).

 

Original Content: 2012-04-29

I am currently a student at Media Design School working on the final component of a 'Post Graduate Diploma of Game Development'. This component is a four month long game project working in a team of fourteen students. The team has elected to create a game based on the intellectual property of a commercial concept artist, we are also using a third party engine for which I had to sign an NDA. Unfortunately due to non disclosure agreements this means I cannot discuss anything relating to the intellectual property in the public arena. It is highly unlikely that this game will be released publically, but it may be used as the base for a future release.

The game is a networked multiplayer third person shooter that is being developed using a commercial game engine. I am building the multi-player framework on top of RakNet. One month in, the basic framework is up and the game can be played at 60fps with 32 players across a local area network. Enough hooks are in place to add prediction and dead reckoning code for Internet play in the future (attempting to test across the Internet creates its own issues proxies and politics). Time will tell if the framerate remains high once final assets (particulary animations) are in place.