Why distinguish UI state from server state?

Question

Accepted Answer

**UI state** is data your app owns and controls locally (toggles, form inputs, selected tab). **Server state** is data that lives on a server and is merely *cached* on the client (fetched users, products, posts). They have fundamentally different characteristics and need different tools.

## The key differences

```text
                  UI state                    Server state
Ownership         you own it (client)         the server owns it; client caches
Freshness         always current              can become STALE (server changed it)
Sync needed?      no                          yes — refetch, invalidate, revalidate
Shared?           usually one client          shared across users/devices/tabs
Concerns          just set/read it            caching, loading, errors, retries, dedup
```

## Why treating server data like UI state goes wrong

```jsx
// ❌ treating server data as plain local state — you must hand-build everything
function Users() {
  const [users, setUsers] = useState([]);
  const [loading, setLoading] = useState(true);
  const [error, setError] = useState(null);

useEffect(() => {
    fetch("/api/users")
      .then(r => r.json()).then(setUsers)
      .catch(setError).finally(() => setLoading(false));
  }, []);
  // and now: caching? refetch on focus? dedup? retries? stale data? all manual 😓
}
```

You end up re-implementing loading/error tracking, caching, refetching, deduplication, and staleness handling — for every fetch.

## The right tool: a server-state library

```jsx
// ✅ React Query / SWR / RTK Query handle caching, loading, errors, refetch for you
function Users() {
  const { data: users, isLoading, error } = useQuery({
    queryKey: ["users"],
    queryFn: () => fetch("/api/users").then(r => r.json()),
  });
  // automatic: caching, background refetch, dedup, retries, stale-while-revalidate
}
```

These libraries are built specifically for the server-state problem — keeping a local cache in sync with a remote source.

## The practical split

```text
UI state    → useState / Zustand / Context / Redux
Server state→ React Query / SWR / RTK Query / Apollo (for GraphQL)
```

## Why it matters

Recognizing that server state is a **cache of remote data** — not regular local state — is a major insight for modern frontend architecture.

It explains why dumping API data into Redux/useState leads to mountains of manual caching/loading code, and why dedicated server-state libraries (React Query, SWR) exist and are so widely adopted.

Splitting the two — UI state in local stores, server state in a caching library — dramatically simplifies data-heavy apps and eliminates a huge category of boilerplate and staleness bugs.