That's nice ... ( hopefully this one will be acceptable )
To paraphrase my previous. The Android EcoSystem is not only Google. Many devices both certified and uncertified access the Google Play Store and A full suite of Google Services ( including Play) are a available to download from reputable 3rd parties given a cursory google search should an end user want to take advantage of them.
jamesh wrote:Android is certainly possible on the Raspi - it just needs a lot of work that the Foundation is not willing to put in, as Android does not have any benefit to the educational side (ie money is better spent elsewhere).
I totally agree and I fully support the foundations line on this. The last thing the world needs is more Java Developers with a specialization on Android App Development.

There are Educational Benefits in Android, however I feel they reside in Actual operating system development rather than anywhere else. Google have dramatically lowered the barrier to entry on this front. They practically made compiling your own OS ( for supported devices ) as easy as typing make.
I suspect this decision by the foundation was arrived at after the ( infamous ) Android 4.0.4 is coming official blog post and as I read elsewhere this effort was very much a personal thing by a single broadcom engineer who is now incognito, Perhaps that post was a mistake and hindsight is wonderful thing eh?
Personally I wouldn't have started my work If I didn't think it was possible or more correctly if I didn't think I was capable of doing it with the resources currently available as I neither expect or want anything from foundation at this point. I would've like it if the Foundations Android forums was less of a den of negativity and ignorance but it's the internet so meh! Plus it's no help when folks show up to a building site without bringing shovels.
Anyway for anyone who cares here's the current state of Android kitkat on the PI
http://pastebin.com/TsQgVixk
As you can see we've initialized the OpenGLES subsystem using the gralloc HAL and bascially the HWComposer isn't getting passed the correct address for the EGLDisplay ATM but is just a case of debugging to find out why. The repeating lines after that is the Bootanimation rendering routines running so I have a feeling this is pretty close ( could be wrong tho )
If anyone should actually fancy a crack with the code then the details are in my previous post ( no not that one, the one before it

)
and like I said to Give up without trying is, well just giving up and that's a poor attitude to have wrt technology ( IMHO )
[Edit : After thought ] @Jamesh Maybe the foundation could have made available details of their decision which possibly would have encouraged some less able/confident folks to get start ... As you don't arrive at a "too much work" decision without researching the matter first.
Thanks
trevd xx
