Putting time outs and other error handling around remote endpoints is prettier much the first lesson you’re taught when interfacing with remote end points.
Yes mistakes happen and I’ve seen other companies ship products that have failed in similar ways, but I’ve always believed Microsoft is thought of as a FANNG and to see a school boy error in their flag ship product a couple of months before it ships really makes me question the standard of leadership running the team.
Who’s QAing it? Why wasn’t those functions covered in unit tests? Who peer reviewed that code? And who’s running the development teams that allow all this guardrails to fail? It’s not just one engineer who had to fuck up to cause that bug.
That fixed release cycle might make certain things easier, but it also causes a forced lag on all changes and encourages shipping code that isn't actually complete so you don't have to tell people to wait another 2 weeks for the change.
I wish some of Kanban was more widely accepted, especially things like making sure you have enough worked planned in the backlog to just keep taking tasks and working on them till completion.
It's ok, we rushed the implementation to meet this sprint's goals, the fix is part of the next sprint. /s