Gear VR – Performance
The processing power and the battery life seem to be extremely limited. Most of the effort will probably go into performance tuning. I should already do the math in the concept stage. There is a rough rule of thumb that says 50.000 – 100.000 quads or vertices are fine.
Of course there should be as few draw calls as possible. I know that from Android experiments that I did a long while ago.
Gear VR – Mobile SDK
I have downloaded the SDK now and have started reading the Mobile SDK Documentation.
I have not actually progressed all the way to the interesting parts, but I know now that I can use Eclipse. That is a plus. I still have to decide how I will develop. I might still go for Unity, after watching an interview. Oculus encourages users to use Unity. A lot of the setup steps sound familiar.
It always feels a little wrong when there are parts in the toolchain that you are not familiar with. The whole Android plugins, toolkits and simulators … those things hurt the overall impression of the development process. I hope I will reach the point where I do most of the work in Eclipse and those things vanish in the background.
At least I was greeted by a friendly message.
Thank you for joining us at the forefront of virtual reality!
That is a lot cooler than the shockingly unfriendly ChallengePost presence.
Gear VR – Content
I am really curious how cool the coolest entries for the challenge will be. I think many studios and people struggle a lot with the platform limits.
On the other hands I am sure there are great applications and games begging to be programmed. I love how this is not just about man power and experience. Creativity is certainly the key to success.
If your angle is getting the idea that you have on the device, chances are you will be disappointed because you need to compromise. The Oculus demo projects for the Unity integration underline this. What they call cool looks pretty plain to anybody … even me and I am easily impressed.
I am trying to find interesting projects on Twitter, Tumblr and Google+. So far nothing really stood out.
The little math I did indicates that a voxel editor might make sense, but a game using voxel sprites would have to be extemely performance optimized by the artist as well as the engine.
I might go for a concept with low poly models and baking textures for detals. Fortunately the challenge motivates me more than the fact that those limits exist drags me down.
For the Mobile VR Jam contest I might do a low poly only game using Unity or the voxel editor using the Native Development Kit.