Project Revitalization news - Aero Empire

I would like to apologize for not explaining anything for so long, and for not releasing the demo game. A lot has happened here at Aero Empire, but the project is not dead, we have kicked into gear, starting a large project revitalization. First, let me explain what happened and the major changes, and then I will show all the progress we have done.

The problems started as we got close to completing the demo game. It's like the 90/10 rule, which states that 50% of your time is spent on the first 90% of the project, and 50% of your time is spent on the last 10% of the project. The final 10% is the integration of everything, which includes debugging, last minute fixes, minor pieces that no one wants to do, and plenty of unexpected problems. We were close, but at the same time far from the completion of the demo, and team motivation was at an all time low. I realized that if this continued, not only would the demo fail, but perhaps even the whole project. So, we started a large project revitalization, to restore team motivation, make changes to the project, and plan things out better so we wouldn't get stuck the same way again.

The first major change is that instead of releasing Aero Empire as one huge insurmountable project all at once, we're going to break it up into small sections, and finish one piece at a time. This way, we can clearly document everything needed for the scope of a single piece, and then we can start releasing games, getting feedback and a sense of completion at the end of each piece. Each piece will cover a single stage, have a small campaign showcasing that stage, and a skirmish mode where you can set up and play custom battles in that stage. This will start with the Gyro stage, and then we will complete the other stages. Each game campaign will have its own complete story focusing on a single event in the Aero Empire world. Eventually, we will release linking programs that link multiple of these individual games together, allowing the player to rise through the stages as was planned for the original Aero Empire - the only difference is that it will be broken into bite sized pieces.

The second major change is that I decided to change Aero Empire's game engine. The Cumulonimbus engine was a failure by my standards, it was the first large game engine I had made from scratch, and so it was not well designed. Some parts of the engine were rushed through, and some parts skipped completely. The engine was just not flexible, not well optimized, and was one of the problems with finishing the game demo. I realized that it would be more work to fix the Cumulonimbus engine than to integrate Aero Empire's effects into a well designed and stable open source engine. After fiddling with several open source game engines, I decided to go with Irrlicht because the API was simple and clean, the program was robust to failure and easy to compile and run on all computers, and even though it didn't have as many features and effects as other game engines, I was going to be porting custom features from the old engine anyways.

One other interesting thing that came about from this switch is that Aero Empire will now run on older graphics cards. We had been thinking for a while that Aero Empire is not really a high end graphics game, and so it was not good that Aero Empire would only run on high end graphics cards. So, as I ported the effects to the Irrlicht engine, I also cleaned them up, optimized them, and added lower quality versions of them that would run on older graphics cards. In higher quality settings, the game should look as good and artistic as it does now. However, there are now settings which will allow the game to run on cards with only OpenGL 1.1 support. To test this, I successfully ran all of the currently ported effects in low quality mode on my Asus Eee Pc which has an Integrated Intel 915GM graphics card and RedOwl's laptop, which has an ATI Mobility Radeon X1400 graphics card. While the game might not look as good depending on your graphics card, it should run on most computers now. Additionally, the cleaning up and optimization made the game run faster on all computers, and there are no longer any problems with the game running on ATI cards (whereas there were some glitches on ATI cards before). It will probably be another two weeks until all the features are ported, but already the results are quite nice even running on old computers.

Aero Empire running on ATI Mobility Radeon X1400 - Running 20 fps at 1440x900 resolution with 4x MSAA, stencil shadows and LDR atmospheric scattering. The model shown has not been updated yet for the new system, which causes the missing triangle problems.

Finally, I am changing the cloud simulation and cloud rendering. I don't think I posted my cloud rant here, but I've been fairly disappointed with Aero Empire's current cloud rendering - while they have some benefits over particle clouds, they are expensive, static and all merge into one gray blob at some angles. So, I already implemented this cloud simulation algorithm: Nis-lab.is.s.u-tokyo.ac.jp , which gives good results and is pretty fast since it uses bitwise operations to run 32 voxel simulations at the same time. I will include several particle cloud rendering techniques for the different quality levels of Aero Empire (and the lowest quality level will work on old graphics cards). Once I finish porting features, I will implement the particle cloud rendering and post some screenshots.

In addition to all of the changes in the project and the game engine, we've had a lot of progress on the game assets, which I would like to showcase here:

In game rendering (old engine) of the preliminary Female Character texture.

Preliminary Male Character model.

In-progress model of the Small Carrier airship.

In-progress model and concept of the Medium Carrier airship.

Model, in game rendering (old engine) and concept of the Skirmisher airship.

Look forward to more updates!

- David