https://github.com/degasus/dolphin/tree/android/Source/Android/app/src
For compiling, you need android studio, at least 2.2 (currently in beta...). And I have to push the fixed gradle file, I'll do so in ~10 hours.
Bad luck, as this is still beta, I can't push it to master. I will once android studio 2.2 is released. This should reduce the barrier to compile dolphin a lot.
By the way, as you likely need some kind of step-by-step assisent, please just ask on IRC: #dolphin-emu @freenode
Sorry to complain... If we had the right stuff, we could prolly hire an Android dev with crowd-funding... eh, yikes, see I kinda don't know how anything like this would be done, cuzz, ya know, I spent my 20s learning guitar instead of anything that like, ya know, helps stuff... anyways... a dream a day!
I doubt the dolphin community will like to buy contributors. We're a voluntary project, and mixing voluntary and paid contributors is a bad idea. Also, I'm sorry, I doubt there is enough money to get paid developers to provide the required "features". ok, input config might work (just a bit of java, there are cheap developers available), but if you're looking for performance improvements, you'll need someone with a lot of C++ and ARMv8 assembly knowledge. Finding one is hard, paying him even harder. 1000$ is just nothing, 10000$ is fine to pay him to understand dolphin and to provide small changes, 100000$ and more if you want to rewrite critical parts of the emulator. Have fun.
I was thinking more about "motivate" those devs that already are working on Android, not hiring them, but I understand your point, Degasus.
Thank you for the responses, It's good to discuss these things.
I feel like motivation is a main factor for anyone who might potentially contribute to Android part of Dolphin. Main android dev who left the project was basically fighting against crap drivers, pushing dozen reports to ARM, Qualcomm and Nvidia, did enormous work maintained on his own. I've been following Android story from v14 Adreno 3xx driver closely, seeing crazy driver regressions and OEMs ignorance. Right now Dolphin contributers tend to ignore Android as there were so many issues with drivers and hardware limitations (S810/808 thermal throttling as example). We need to see how it'll perform on S820 and QC ES 3.2 drivers (Exynos 8890 is badly messed this time + no buffer_storage), depending on the results it may/may not get devs interested to put their efforts working on Android port. Vulkan backend too. Not so many people wanting to invest their time in something so vague, you need a curiosity for that or you need to know that it'll just work.
zxcvbad: Yes, you've got the point. And it's worse. Most devs want to improve the emulator for themself, so they are only working on the plattform they use on their own. That's also why we have so big problems on OSX right now. So compared to android, were no device is able to run as fast as you can enjoy many games, no one wants to improve it to actually use it. That's a big point for the motivation.
I'm wondering how many "nvidia shield TV" have been sold... I mean, Success or fail? It seems like we need to have a lot of people with the console for real changes... sadly shield TV is not a 3DS.
I understand the situation, but is very sad because, perfect for dolphin or not, Shield TV is for me an awesome option for emulation and alternative gaming in the living room/TV, specially in multiplayer local games (better than in front of a PC).
Well, it's just about time. Patient, patient
It's by far the best streaming device available. How many mobiles have been sold? But there is still almost nobody porting dolphin to mobiles
And don't expect your shield to be able to play dolphin will. Right now, it's the best android device for dolphin. But the performance is by far to bad to play lots of games. Don't think this is because we lack manpower and i will increase a lot. Through there are always ways to optimize it better, we're already quite good. So "big" improvements are like 1-5% performance. The shield is still a 10W device, shared with a powerful GPU. The CPU is just slow.
How do you think the SD 830 will compare to the shield TV? Apparently they're gonna do at least some work on sustained performance, and with vulkan and buffer storage, it may do well. And we can always hope Qualcomm decide to magically improve their drivers in that time.
Looks like thermal throttling is not an issue on S820, when thermal throttling it gets you a similar performance as peak performance on S810 + better GPU (A57s utilization). Simple profiling exposing it easily