I started a similar thread to this on my forum, but thought I would get everyone's perspective here as well. I've been working on an Android port of Mupen64Plus (an open-source Nintendo 64 emulator) for the past couple years, and have published a Beta version on Google Play. It is still pretty rough and definitely a work in progress, but lots of games are playable at this point. Would there be any interest in me porting this to the OUYA? Specifically, would there be a lot of people willing to test the emulator, and developers willing to work on the source code?

Now the first obvious consideration is that the GUI would need significant changes, since this is not a touch-screen device, but rather it is interfaced like any console, through game controllers. Supporting an interface like this might be a good thing even for the Android project, though, because Android is installed on all types of hardware (and this diversity will only continue to expand in the future).

Secondly, I haven't confirmed if the OUYA SDK also has an NDK counterpart, or if it can link with libraries built with the Android NDK (I would assume one or both of these are possible, but need confirmation). This of course would be essential, since the emulator itself is written in native C/ C++.

I also don't know if the OUYA is built with GLES2 (probably) or normal OpenGL (we do have plug-ins for both, though). I know the CPU is the Tegra 3, so we'd be using Ari64's dynarec of course, just like the Android port.

One final issue is I doubt the OUYA comes with a native SDL library installed (and the Android port would most likely not work out-of-the-box for it). On the other hand, I am going to remove the SDL requirement from Mupen64Plus AE, so this isn't really a game-ender like it might sound at first.

So what does everyone think? Would adding support for the OUYA be a good idea at this point, or should our focus be only on improving the Android version we have now to eventually get it to a level of compatibility where we can say we've officially progressed past the Beta testing phase? One thought I had is that having support for OUYA built into our source code repository might bring new developers from the OUYA community on board which would have a positive impact on the emulator overall (but maybe that's just wishful thinking on my part!)