ubiquiti: cite both mobile-router rows and give the UMR its 10 W passthrough budget #2128
No reviewers
Labels
No milestone
No project
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference
jaylouisw/packetforge!2128
Loading…
Add table
Add a link
Reference in a new issue
No description provided.
Delete branch "catalogue/1119-0922"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Refs #130 — does not close it; the Dream Wall Pro is still absent from Ubiquiti's catalogue.
What was wrong
devices/ubiquiti/umr.jsonanddevices/ubiquiti/umr-ultra.json— the two cellular gateways that answer this request's stated lab need (routing, NAT, DHCP, cellular WAN) — carried nosourceUrlat all: theirnotesquoted vendor figures with no URL behind them. On top of that the UMR's PoE passthrough port had nopoe.budgetW, andengine/simulator.jsopens its allocator withif (!dev.poeBudget) continue;, so the port was skipped rather than reported unknown: it could power nothing andshow power inlinecould not say why.Evidence, read 2026-11-19
techspecs.ui.com/unifi/integrations/umr-usdl.ui.com/ds/umr_ds.pdfapplication/pdf, 597,924 B, 2 ppstore.ui.com/us/en/products/umr-ustechspecs.ui.com/unifi/integrations/umr-ultra-usstore.ui.com/us/en/products/umr-ultra-usThe UMR's own two documents disagree on the chassis only (datasheet 220.6 × 89.8 × 28.8 mm, 252/298 g against tech-spec 190 × 89.8 × 28.8 mm, 257/303 g). Ports, the 150 Mbps WiFi 4 rate and the 9 W draw are identical on both, so the row models those and nothing from the disagreement.
Changes
umr.json— cites all three surfaces;poe.budgetW: 10(vendor's "Max. 10W", on the tech-spec page only);poeInis 9 W instead oftrue(docs/DEVICE_SCHEMA.md:204wants watts); capabilities gainpoe-pd,poe-pse,802.3at; aliases carryUMR-US.umr-ultra.json— cites both surfaces; aliases carryUMR-Ultra-US.research/ubiquiti-mobile-router-census.md— the census table and the retrieval traps.Effect on the box, cabled to a 12.95 W 802.3af camera:
The 10 W ceiling is the vendor's, so refusing a 12.95 W class-0 device is the correct answer; what was missing was the number, not the policy.
Traps recorded so nobody re-walks them
-ussuffix is part of the slug on both hosts:techspecs.ui.com/unifi/integrations/umranswers the 609,911-byte front page (the silent fallback), andstore.ui.com/us/en/products/umris a 404./products/umr-ultra307s to the internet-solutions category — a 200 once followed, but a miss, not a product page.dl.ui.com/ds/umr-ultra_ds.pdfand.../umr-ultra-us_ds.pdfboth return the 609,911-byte techspecs front page.rj45-1sinks andrj45-2sources on the vendor's stated count, thezyxel/gs1900-24ep.jsonconvention.Gates
node scripts/syntax-check.js→ 446 files, all OKnode scripts/audit-port-types.js→ "No port contradicts its declared type."node scripts/check-control-bytes.js→ OKnode scripts/import-devices.js --dry-run→ noumr/umr-ultradiagnostics (its two rejects are the pre-existing peplink/zte duplicates inresearch/incoming/)node test/hardware-invariants.test.js→ 26 passed, 0 failednode test/poe-inference.test.js,test/poe.test.js,test/device-aliases.test.js,test/capability-vocabulary.test.js→ 17 / 18 / 13 / 13 passednode scripts/run-tests.js→ 367 suites, 12,468 assertions, 0 failednpm run regen:registry-doc→ "The catalogue pages are current: 2,356 models"