The concerns, constraints, and behavior of server-side headless software and a viewer that fetches and renders data from a remote backing store are entirely different.
Is it a bad thing that organizations specialize on those concerns? Seems like the most obvious place to slice responsibility, even if it means every feature is split across two teams. After all, the performance of the backing store is impacted not by neigbhoring logic in the UI infrastructure but by neighboring state in the backing store; it should be someone's job to be looking at the backing infrastructure holistically and not sliced as "frontend-backend feature A, frontend-backend feature B, etc."
Is it a bad thing that organizations specialize on those concerns? Seems like the most obvious place to slice responsibility, even if it means every feature is split across two teams. After all, the performance of the backing store is impacted not by neigbhoring logic in the UI infrastructure but by neighboring state in the backing store; it should be someone's job to be looking at the backing infrastructure holistically and not sliced as "frontend-backend feature A, frontend-backend feature B, etc."