docs(device-registry): generate the open-request section the issue threads cite #1847

Merged
jaylouisw merged 1 commit from catalogue/1117-1255 into main 2026-11-17 13:05:33 +00:00
Owner

Refs #130, #742, #743, #803, #807, #808, #810 - repairs the record they cite; closes none of them.

The defect. The blocked-[Device] threads point at docs/DEVICE_REGISTRY.md section "What is left open, and why" as the place where each refusal is argued. Grep that page and the string is absent: the registry only ever wrote catalogue totals, so five separate threads cite a section that does not exist.

The fix. The blocker data moves to research/open-device-requests.json and the section is rendered by scripts/regen-registry-doc.js between <!-- open-requests:start --> / end markers, alongside the totals that page already regenerates. Adding or closing a request is one JSON edit plus npm run regen:registry-doc; nothing is hand-written into a generated page.

Each row states the field the vendor does not publish, what would unblock it, the date last checked and the URLs read - all carried over from the threads that did that work. No new vendor claim, no new spec, no engine change.

Queue state at 2026-11-17 12:55Z (all seven requests, re-verified this pass):

issue blocked on
#130 the SKU udw-pro does not exist in Ubiquiti's own gateway index, datasheet or store
#742 front-end port count per node - data sheet stops at port types
#743 per-model IOM slot rules + module port counts; both books are served section-by-section and their slugs are unindexed
#803 no board list for the E6600 generation outside the closed hedex tree
#807 / #808 / #810 EDOC1100163901 board datasheet is vendor-linked and 404 on every anonymous route

Checked this pass and found nothing new: support.huawei.com product pages under a crawler UA (the X-series page exposes a Downloadable Product Documentation Package card -> doc-set EDOC1100368651, already recorded in research/oceanprotect-x-series-port-census.md section 11); the same idPath query the product page itself uses (302 -> /view/notFound); Chrome and curl user-agents against info.support.huawei.com/enterprise/en/doc/EDOC1100305200/<slug> (both 302, X-ALB-HIT: req-hd-bl, so the rule is not UA-keyed); and search-index probes for the X8100/X9100 books, whose section pages are not indexed while the Dorado and 1.5.0 books' are.

Gates: scripts/syntax-check.js 386 files OK; scripts/audit-port-types.js 0 offending devices / 0 offending ports; scripts/check-control-bytes.js clean; test/doc-catalog-counts.test.js 43/43 including the staleness paths; node scripts/regen-registry-doc.js --check current at 2,348 models / 258 vendors / 29,235 sockets / 966 shapes.

Refs #130, #742, #743, #803, #807, #808, #810 - repairs the record they cite; closes none of them. **The defect.** The blocked-[Device] threads point at `docs/DEVICE_REGISTRY.md` section *"What is left open, and why"* as the place where each refusal is argued. Grep that page and the string is absent: the registry only ever wrote catalogue totals, so five separate threads cite a section that does not exist. **The fix.** The blocker data moves to `research/open-device-requests.json` and the section is rendered by `scripts/regen-registry-doc.js` between `<!-- open-requests:start -->` / `end` markers, alongside the totals that page already regenerates. Adding or closing a request is one JSON edit plus `npm run regen:registry-doc`; nothing is hand-written into a generated page. Each row states the field the vendor does not publish, what would unblock it, the date last checked and the URLs read - all carried over from the threads that did that work. No new vendor claim, no new spec, no engine change. **Queue state at 2026-11-17 12:55Z** (all seven requests, re-verified this pass): | issue | blocked on | |---:|---| | #130 | the SKU `udw-pro` does not exist in Ubiquiti's own gateway index, datasheet or store | | #742 | front-end port count per node - data sheet stops at port *types* | | #743 | per-model IOM slot rules + module port counts; both books are served section-by-section and their slugs are unindexed | | #803 | no board list for the E6600 generation outside the closed hedex tree | | #807 / #808 / #810 | `EDOC1100163901` board datasheet is vendor-linked and 404 on every anonymous route | **Checked this pass and found nothing new:** `support.huawei.com` product pages under a crawler UA (the X-series page exposes a `Downloadable Product Documentation Package` card -> doc-set `EDOC1100368651`, already recorded in `research/oceanprotect-x-series-port-census.md` section 11); the same `idPath` query the product page itself uses (`302 -> /view/notFound`); Chrome and curl user-agents against `info.support.huawei.com/enterprise/en/doc/EDOC1100305200/<slug>` (both `302`, `X-ALB-HIT: req-hd-bl`, so the rule is not UA-keyed); and search-index probes for the X8100/X9100 books, whose section pages are not indexed while the Dorado and 1.5.0 books' are. **Gates:** `scripts/syntax-check.js` 386 files OK; `scripts/audit-port-types.js` 0 offending devices / 0 offending ports; `scripts/check-control-bytes.js` clean; `test/doc-catalog-counts.test.js` 43/43 including the staleness paths; `node scripts/regen-registry-doc.js --check` current at 2,348 models / 258 vendors / 29,235 sockets / 966 shapes.
docs(device-registry): generate the open-request section the issue threads cite
All checks were successful
CI / Classify the change (pull_request) Successful in 8s
PR rebase / Rebase onto main and arm auto-merge (pull_request) Successful in 7s
CI / Server smoke test (pull_request) Successful in 8s
CI / Docker build (pull_request) Successful in 22s
CI / Test (Node 18) (pull_request) Successful in 54s
CI / Test (Node 20) (pull_request) Successful in 48s
CI / Test (Node 22) (pull_request) Successful in 50s
CI / Arm auto-merge (pull_request) Successful in 24s
3463112da0
Five blocked-[Device] threads name docs/DEVICE_REGISTRY.md section "What is left
open, and why" as the record for why each request cannot be modelled. That
section did not exist: the registry page only ever wrote catalogue totals, so
every one of those citations pointed at nothing.

The data now lives in research/open-device-requests.json and the section is
rendered by scripts/regen-registry-doc.js between markers, next to the other
regenerated totals - so a request is added or closed with one JSON edit plus
npm run regen:registry-doc, and no hand-edited prose about a directory.

Fields are the ones already measured in the threads: the missing field, what
would unblock it, the last-checked date and the URLs read. No new vendor claim
is made and no device spec is added.
Sign in to join this conversation.
No reviewers
No milestone
No project
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference
jaylouisw/packetforge!1847
No description provided.