Recently I discovered that, when implementing device orientation detection methods, android devices return acceleration-gravity vectors that are the exact opposite of the iOS-safari implementation. This was not covered in existing literature, although it may be related to the fact that rotation sign has to be reversed in iOS webkit.
It’s annoying that the vendors can’t all agree with the standard, and we will be left with many window.chrome? window.safari? hacks to guess the device used.
I might as well get a reference android tablet to test apps in the future. Recently, the nexus 7 (2013) can be obtained for ~ $100 online, and can be a useful testing device. That being said I still dislike the overall feel and design of android. Just doesn’t feel right, and doesn’t feel polished.