Changing WiFi stack is less of a problem than changing the implementation language. The perceived fragility of a an element of architecture should be inversely

Changing WiFi stack is less of a problem than changing the implementation language. The perceived fragility of a an element of architecture should be inversely proportional to the effort of replacing the said module chosen. Large effort to replace an element (like the implementation languag) means the element should be of low fragility. This is not an argument against Zig or that it would not become an industrial scale tool. It just does not feel "lindyproof" yet for real work.