patches broke for CM11-M6

[Lanchon]

this commit broke compatibility:

CyanogenMod/android_frameworks_base@fe9b529

by changing CONNECTIVITY_SERVICE from a static service to a context dependent one.

mateor, are you still working on this? would you like to patch this up or should somebody else do it? also, there is code like this:

                // BEGIN privacy-modified
                //return new LocationManager(ctx, ILocationManager.Stub.asInterface(b));
                Log.d(TAG, "PDroid:ContextImpl: returning PrivacyLocationManager");
                // SM: I'm not sure whyt this is using getStaticOuterContext rather than getOuterContext.
                // Would have thought it should have been the following line:
                // return new PrivacyLocationManager(ILocationManager.Stub.asInterface(b), ctx.getOuterContext());
                return new PrivacyLocationManager(ILocationManager.Stub.asInterface(b), getStaticOuterContext());
                // END privacy-modified                    

this led me to look a bit into the code. it seems that ContextImpl.getStaticOuterContext() should always be == ctx since there is exactly one ContextImpl instance per process. is this true? and is ctx.getOuterContext() not always == ctx because somebody calls setOuterContext? it would seem so, otherwise what would be the purpose of xxxOuterContext? can you shed some light on this?

getStaticOuterContext is potentially unsafe. is it me or is OPDs codebase kind of dirty?

[mateor]

The confusion you see over the outerContext is remnant of the CollegeDev
/ FFU5y work. CollegeDev would drop huge chunks of code all at once,
some of which would end up in OpenPDroid. But it often took lots of
vetting and editing, and there are plenty of parts where I didn't merge
his work simply because I didn't have the time to try and see the
reasoning behind it. We all could have benefited from a better workflow
and better communication.

The code base needs tons of attention. I was confident that it was
(mostly) performing its function, but as time went on, I became less
confident. Since I do not have the time to give or organize a full
audit, I chose to end the project from my end. Better to be upfront.

I will take a quick look at the breakage. I am willing to very lightly
maintain the actual source code patches. But I have some pretty
significant open-source commitments elsewhere, so we will just have to
see what this breakage entails. I will let you know the status in a day
or two.

[Lanchon]

never mind, you seem to be too busy, i'll do it and push you the changes.

what are you working on now?

[Lanchon]

ive fix this and some other stuff, while i compile android ill write about it

[mateor]

I working on a couple projects, but the open-source project is a build
system. The build system supports things like Java, Python, Thrift, etc.
I extending it to support Android applications. It is a competing
product to something like gradle, although still very much in the
creation stage.

I put a brief note on your commit and will read your write-up tonight,
thanks for contributing!