That's not the first story like this. It seems like more and more often the QA is done by users who have not signed up for it. Why can't a multibillion company do a proper QA?
I wouldn't be surprised if it ends up making them money: people will replace their devices with newer ones. Once the device is sold it no longer has a function for the manufacturer until the next sale happens to the same user. So any update that bricks a percentage of the devices (or makes them effectively unusable) may well extract some more $ from the users.
Sure, but this is the past half dozen models of the phone made by the same company who makes the OS. You'd think these would be the easiest possible test cases for them to do up front.
I assume it's just a fraction of users affected, so it could be something like a service reading third-party app data and crashing due to a bug. The kind of thing that should be caught in code review, because comprehensive testing is next to impossible.
Because most test devices are likely constantly factory reset and aren't given the chance to live with multiple user profiles on them for a long period of time where cruft can build up and people can notice that using using external storage is broken.
Staffing proper QA teams went out of vogue a couple of years ago.
Turns out that the way incentives are set up at most tech companies today, nobody gets dinged for shipping major software regressions/bugs while everyone is patted on the back for shipping even completely broken features on time.
The fact that they have the ability to roll this out to a handful of users, all of whose devices stopped working, and to detect that and not proceed with the rollout is considered success, by them.
