I do have some apps crash, but that's the app developer's problem. Not much the OS vendor can do about that.
I've written a somewhat popular Android app and most of its crashes are either the fault of the runtime or of vendor specific customizations.
The Developer Console provides excellent reporting on both uncaught Java exceptions and crashes in native code. Most often, the Dalvik VM crashes during garbage collection. The Dalvik class loader is also flaky and has issues with multithreading that only got worked around in newer versions. I've also seen the libchromium crash inside my app from adverts delivered by Googles Admob service, which is somewhat scary from a security perspective.
Another issue is that apps using Googles compatibility library now crash on HTC devices running Android 4.1 as soon as the user presses the menu key. HTC is no longer providing updates and Google simply states that it's not their fault and therefore not their problem. Now thousands of app developers have to independently find out about the issue and work around it.
So when an app crashes, there's a good chance that it's not the fault of the app developer. On a brighter note, the ART VM used in Android 5 and newer seem to be rock solid.
Videoconferencing from any device on the planet without installing any special software is bloat?
YES, in the same way that every user on the planet would probably want a calculator once in a while but that doesn't mean the browser needs to add one!
Firefox comes with a couple of calculators built in. It has since before it was called Firefox.
Usually, "epoxy" around the edges of a BGA chip is neither an anti-hacking attempt nor a light-proofing attempt. It's called underfill, and its chief purpose is to increase mechanical strength and make the bond more durable than tiny bare solder balls would be on their own.
Yes they are. Most multimedia processing is parallelizable, and thus benefits greatly from SIMD instructions - for example, just about every CPU-based video codec ever. If you want an actual example, I wrote a high-performance edge detection algorithm for laser tracing, with its convolution cores written in optimized in SSE2 assembly, and am hoping to write a NEON version. It'll never run reasonably on the original Raspberry Pi because it's too underpowered to do it without SIMD (I didn't even bother writing a plain C version of the cores, because honestly any platforms without SSE2 or NEON are going to be too slow to use anyway).
Obviously you can use SIMD instructions for a lot more, but multimedia is the obvious example. And as I mentioned, the Pi makes up for it for standard codecs only with its GPU blob decoder, but that doesn't help you with anything that isn't video decoding (e.g. filtering).
8 Catfish = 1 Octo-puss