[fpgeek]

> False. The software running on your Android phone's application processor is not open source.

You're skipping over an important practical distinction: The precise software running on the application processor may not be open source, but it is closely related to usable software that is. Given that, it's possible to learn quite a bit by comparing the behavior of the closed-source fork and the open-source base.

Is that as good as "open source all the way down"? Of course not. But it is a hell of a lot better than the "opaque binary blobs all the way down" offered by most of the alternatives.

[coldtea]

>Is that as good as "open source all the way down"? Of course not. But it is a hell of a lot better than the "opaque binary blobs all the way down" offered by most of the alternatives.

I fail to see how. It's not like a partial binary blob is better than a full on binary blob. The opaque part might do anything too...

[fpgeek]

Well, consider the example of Carrier IQ.

The Carrier IQ software was installed on some Android phones, some iPhones and some Blackberry devices: http://www.computerworld.com/s/article/print/9222319/AT_T_Sp...

Where was Carrier IQ found first? Why?

[coldtea]

>Where was Carrier IQ found first? Why?

The link doesn't say. And if anything it was not because there was an open-source part of the phone OS.

For one, on active, sold, phones, the device code is compiled anyway.

[gress]

I'm not skipping anything, and at least you are agreeing that shipping Android phones are not using open source software.

I do not dispute that having a related system that is open-source can aid in reverse engineering.

However the source of commercial Android phones is not open source, and does not have the benefits that open source would imply.