Ionosphere.tv
perf: exclude document column from getTalks response
getTalks was returning 96MB because it included the full faceted
transcript document for every talk. The document column is only needed
for individual talk views (getTalk). This was causing build OOMs.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
+6
-1
+6
-1
apps/ionosphere-appview/src/routes.ts
+6
-1
apps/ionosphere-appview/src/routes.ts
···
23
23
app.get("/health", (c) => c.json({ status: "ok" }));
24
24
25
25
app.get("/xrpc/tv.ionosphere.getTalks", (c) => {
26
+
// Exclude 'document' column — it's the full faceted transcript (up to 2MB per talk)
27
+
// and not needed for list views. Individual talks are fetched via getTalk.
26
28
const talks = db
27
29
.prepare(
28
-
`SELECT t.*, GROUP_CONCAT(s.name) as speaker_names
30
+
`SELECT t.uri, t.did, t.rkey, t.title, t.description, t.video_uri,
31
+
t.video_offset_ns, t.video_segments, t.schedule_uri, t.event_uri,
32
+
t.room, t.category, t.talk_type, t.starts_at, t.ends_at, t.duration,
33
+
t.created_at, GROUP_CONCAT(s.name) as speaker_names
29
34
FROM talks t
30
35
LEFT JOIN talk_speakers ts ON t.uri = ts.talk_uri
31
36
LEFT JOIN speakers s ON ts.speaker_uri = s.uri