fix: early sentinel trigger + short author prefetch timeout

- Feed sentinel now fires 400px before the bottom (rootMargin: '400px')
  so the next page fetch starts while the user is still reading, hiding
  the latency entirely in most cases

- Author prefetch timeout cut from 5000ms to 1500ms — relay responds
  well within that window for cached/known profiles; cards whose authors
  miss the deadline fall back to individual useAuthor queries rather than
  blocking the whole page render

- Stats prefetch reverted to fire-and-forget — not worth blocking on,
  useEventStats per-card handles misses with its own skeleton state

Co-authored-by: shakespeare.diy <assistant@shakespeare.diy>

src/components/Feed.tsx

@@ -42,7 +42,7 @@ export function Feed() {
     await queryClient.invalidateQueries({ queryKey: ['feed', activeTab] });
   }, [queryClient, activeTab]);
 
-  const { ref, inView } = useInView();
+  const { ref, inView } = useInView({ rootMargin: '400px' });
 
   useEffect(() => {
     if (inView && hasNextPage && !isFetching) {

src/hooks/useFeed.ts

@@ -71,34 +71,36 @@ async function prefetchPageData(
     .map((item) => item.event.id)
     .filter((id) => queryClient.getQueryData(['event-stats', id]) === undefined);
 
-  await Promise.all([
-    pubkeysToFetch.length > 0
-      ? nostr.query(
-          [{ kinds: [0], authors: pubkeysToFetch }],
-          { signal: AbortSignal.timeout(3000) },
-        ).then((profileEvents) => {
-          for (const ev of profileEvents) {
-            let metadata: NostrMetadata | undefined;
-            try { metadata = n.json().pipe(n.metadata()).parse(ev.content); } catch { /* skip */ }
-            seedAuthorCache(queryClient, ev.pubkey, { event: ev, metadata });
-          }
-        }).catch(() => {})
-      : Promise.resolve(),
+  // Await author prefetch with a short deadline — whatever profiles arrive
+  // in time get cached; cards whose authors miss the window fall back to
+  // individual useAuthor queries (acceptable for a minority of cards).
+  if (pubkeysToFetch.length > 0) {
+    await nostr.query(
+      [{ kinds: [0], authors: pubkeysToFetch }],
+      { signal: AbortSignal.timeout(1500) },
+    ).then((profileEvents) => {
+      for (const ev of profileEvents) {
+        let metadata: NostrMetadata | undefined;
+        try { metadata = n.json().pipe(n.metadata()).parse(ev.content); } catch { /* skip */ }
+        seedAuthorCache(queryClient, ev.pubkey, { event: ev, metadata });
+      }
+    }).catch(() => {});
+  }
 
-    eventIdsToFetch.length > 0
-      ? nostr.query(
-          [
-            { kinds: [1, 6, 7, 9735], '#e': eventIdsToFetch, limit: eventIdsToFetch.length * 10 },
-            { kinds: [1], '#q': eventIdsToFetch, limit: eventIdsToFetch.length * 3 },
-          ],
-          { signal: AbortSignal.timeout(6000) },
-        ).then((statEvents) => {
-          for (const id of eventIdsToFetch) {
-            queryClient.setQueryData(['event-stats', id], computePageStats(id, statEvents));
-          }
-        }).catch(() => {})
-      : Promise.resolve(),
-  ]);
+  // Stats are fire-and-forget — nice to have pre-cached but not worth blocking on.
+  if (eventIdsToFetch.length > 0) {
+    nostr.query(
+      [
+        { kinds: [1, 6, 7, 9735], '#e': eventIdsToFetch, limit: eventIdsToFetch.length * 10 },
+        { kinds: [1], '#q': eventIdsToFetch, limit: eventIdsToFetch.length * 3 },
+      ],
+      { signal: AbortSignal.timeout(6000) },
+    ).then((statEvents) => {
+      for (const id of eventIdsToFetch) {
+        queryClient.setQueryData(['event-stats', id], computePageStats(id, statEvents));
+      }
+    }).catch(() => {});
+  }
 }
 
 /** Hook to fetch the global or followed feed with infinite scroll pagination. */