Probably not “unsolvable” but I imagine not well documented and a million footguns to make RN work in the scenarios described in the article. This would entail

Probably not “unsolvable” but I imagine not well documented and a million footguns to make RN work in the scenarios described in the article. This would entail working in a new fork of React Native to add implementations for the underlying Android and iOS SDKs where the native functionality exists. That might include language bindings at multiple levels, e.g. from JS to Java, and potentially Java to C if NDK is required (and that’s just for android) In my experience, writing those bindings is very bug-prone and user-friendly documentation can be hard to come by. We wrote some Python extensions in C for a very simple functionality where we wanted native speed. Could definitely see it having been easier to just write the dang program in C, since you had to understand C as well as some fundamentals about how the Python runtime works in C along with its special data types and execution patterns. And if they mainly hire app developers and not interpreter/OS developers that could become entirely untenable