Area: docs — accuracy · found via docs review on #522
Three published pages point readers at closed issues for work that is either unfinished, retracked elsewhere, or explicitly not planned. Same defect class as #509 and #523: prose that outlived its subject and now hands the reader a dead tracker.
All three are in files #522 doesn't touch, so they're filed rather than folded in — #522 is already carrying two out-of-scope corrections.
1. docs/src/content/docs/durability.md:90 — promises a CLI that isn't tracked
"A self-contained wavehouse storage-check preflight subcommand that bakes this measurement and verdict into the binary … is tracked in #84."
Fix: either file the feature issue and re-point, or reword to say plainly that it isn't implemented and isn't currently tracked. Right now the page promises a preflight command that nobody is working on.
2. docs/src/content/docs/ingest-pipeline.md:258 — a whole section tracked to a shipped refactor
The ## Deferred / not yet implemented section is headed "Tracked under #191:", but #191 is CLOSED/COMPLETED — it's the refactor that shipped. The four bullets are still substantively accurate; only the tracking is wrong, in three different ways:
| Bullet |
Real tracker |
tableLoop reaping |
#263 (open, near-verbatim title) |
| sweeper / clustering |
#264 (open) |
| pipelining beyond coalescing |
none |
| per-table / partitioned consumers + two-stream retention |
none |
(bullet 4's #214) |
still live ✅ |
Also fix the code side: internal/ingest/worker.go:223 carries TODO(#191): tableLoops are spawned per distinct table and never … — the same dead pointer. Fixing the doc alone leaves half the drift.
Fix: re-point per bullet (#263, #264), and file or drop the two untracked ones.
3. docs/src/content/docs/sdk/streaming.md:115 and :188 — cites an issue closed as not planned
Both cite #449 for the live-query backfill-seam dedup defect. #449 was closed NOT_PLANNED on 2026-08-25 with: "Duplicate of #396. Same bug, same lines … Everything unique to this issue has been ported to #396."
#396 (bug(sdk): liveQuery dedup boundary is wrong — duplicate and dropped events in the core streaming feature) is open and is the superset.
The surrounding prose is accurate — only the pointer is dead. But "not planned" is the worst possible signal to send a reader about a live bug in a core streaming feature.
Fix: swap both references to #396.
Related: #523 (same review pass, OTLP port claim), #521 / #522 (the WH001 work that surfaced these)
Area: docs — accuracy · found via docs review on #522
Three published pages point readers at closed issues for work that is either unfinished, retracked elsewhere, or explicitly not planned. Same defect class as #509 and #523: prose that outlived its subject and now hands the reader a dead tracker.
All three are in files #522 doesn't touch, so they're filed rather than folded in — #522 is already carrying two out-of-scope corrections.
1.
docs/src/content/docs/durability.md:90— promises a CLI that isn't trackedCLOSED/COMPLETED, and was never a feature issue — it's titled "docs(deploy): JetStream SyncAlways=true durability vs throughput on commodity storage". Its own comment says "mq: expose JetStream sync_interval as a config knob (throughput vs durability tradeoff) #139 = code-side knob; docs(deploy): JetStream SyncAlways=true durability vs throughput on commodity storage #84 = 'when and why' docs". It was closed by writing that page.cmd/wavehouse/contains onlymain.go,health.go,validate.go— there is nostorage-check, and no open issue tracks one.Fix: either file the feature issue and re-point, or reword to say plainly that it isn't implemented and isn't currently tracked. Right now the page promises a preflight command that nobody is working on.
2.
docs/src/content/docs/ingest-pipeline.md:258— a whole section tracked to a shipped refactorThe
## Deferred / not yet implementedsection is headed "Tracked under #191:", but #191 isCLOSED/COMPLETED— it's the refactor that shipped. The four bullets are still substantively accurate; only the tracking is wrong, in three different ways:tableLoopreaping#214)Also fix the code side:
internal/ingest/worker.go:223carriesTODO(#191): tableLoops are spawned per distinct table and never …— the same dead pointer. Fixing the doc alone leaves half the drift.Fix: re-point per bullet (#263, #264), and file or drop the two untracked ones.
3.
docs/src/content/docs/sdk/streaming.md:115and:188— cites an issue closed as not plannedBoth cite #449 for the live-query backfill-seam dedup defect. #449 was closed
NOT_PLANNEDon 2026-08-25 with: "Duplicate of #396. Same bug, same lines … Everything unique to this issue has been ported to #396."#396 (
bug(sdk): liveQuery dedup boundary is wrong — duplicate and dropped events in the core streaming feature) is open and is the superset.The surrounding prose is accurate — only the pointer is dead. But "not planned" is the worst possible signal to send a reader about a live bug in a core streaming feature.
Fix: swap both references to #396.
Related: #523 (same review pass, OTLP port claim), #521 / #522 (the WH001 work that surfaced these)