as seen on d-visions.com

Several months ago, while freelancing at EVB in San Francisco, I met an intersting young man by the name of Ben McChesney. He attended RIT, not far from my native city of Buffalo. Whilst working on this project we discussed the powers of Unity and he turned me onto how easy it is to get Kinect working in the IDE. When I went back home to NYC, I went out and bought myself a Kinect. Purely for development purposes, as I don’t own an Xbox. Within just a couple hours I was up running and tracking my avatar in Unity. I had taken my first step into a larger world.

Since then I have attended some OpenKinect events here in NYC. One hosted by M$ that made me realize they have no clue how to market this technology. Another with PrimeSense, the company who developed the underlying technology for the Kinect to begin with. At this event I was able to see first hand the capabilities demonstrated by the great Shlomo Zippel. Many of which I had seen as they are the demos included with the drivers. But he keyed me into what was being done with Unity. For months I had been visiting a gitHub only to come up with no new demos/updates. Now it seems the drivers come with full fledged examples. I was trying to hack my way around with the initial gitHub files and only got so far. Prior to this event I had an epiphany.

So this all leads to my demo. I wanted to capture a simple tossing motion like that of a frisbee for the throwing star. I assumed such a gesture would be canned into the drivers/middleware themselves, I was wrong. As has become the norm lately you need to roll your own. M$ alluded to this at their event and was something they were trying to correct with their SDK. Now prior to either of these events I had already been working on this with a simple verlet integrator. Check velocity over a few frames and if force is greater than threshold release. Problem solved…

Well the Adobe MAX keynote came and went. After dragging out the presentation thru 4 moderate areas of interest, they finally got to what I really care about, Games. They spent only about 10-15min showing what we may be able to do someday and everybody ejaculated all over themselves, but I believe that to be premature.

Here is what they didn’t tell you, but I can read thru the lines. First, as with most things Flash this will still require a 3rd party or OS project to really render your worlds. Of course you could write you own engine, but isn’t that what we expected this next-generation API to do for us? Second, they didn’t even hint at a tool. Have you ever done 3D? Ever try to do it all in your head? Good luck! Isn’t it easier to have a tool to see visually whats happening and edit on the fly? Director was always like this, Unity too. But Flash we always have to compile down, test, tweak something and recompile – repeat this tired loop until you get it just right. Sure there are 3rd party tools to do this at runtime and other methods for such manipulation, but shouldn’t our tool do this for us? Seb-Lee Delisle recently wrote an article Flash vs Flash about how the tool conflicts with some initiatives of the platform. I always pondered similar things such as why anyone would use youtube to convert a swf animation to video at all? I digress.

Again, what should be taken away from this grand announcement is it won’t be available soon. They say first half of of 2011 for a beta program. I’d say don’t expect it long before June 30th. Remember when they announced at MAX 2009 iphone packager was going to be a public beta by the end of last year? Expect similar treatment. They’ve damn near guaranteed it will be difficult to use as it is just a low-level API and is NOT an engine. Oh yea go watch this video in which he say the physics are ActionScripted. WTF? Are you telling me along with this great new API that I have to code an engine in, I also get to code 3D physics to make my world work? Wow! Thanks Adobe for just tripling my workload before I even begin my game.

If 3D in Flash is a molehill, then Unity is my mountain in which I will stay atop.