> Angular’s change detection is a disaster. The developer gets two suboptimal choices: (1) the slow and naive default implementation, or the complexity of manag

> Angular’s change detection is a disaster. The developer gets two suboptimal choices: (1) the slow and naive default implementation, or the complexity of managing change detection manually. This is completely wrong. The "naive" approach is the one you should always use, with the onpush strategy reserved for breaking certain cascading situations manually. But the default approach works perfectly fine, it is performant enough, and even more when compiled for production, as it runs only once (as compared to twice in the debug mode). The author is incorrectly assuming that the change detection is slow because it is automatic. That's the entire point of the framework btw.