[AidenJohnston]

What changed

Adds experiments/device-field-test/build-4.1/, which holds the analyses of the three TestFlight 4.1 recordings from 2026-09-26:

• Run 1 (20:00): the mock wall, indoors.
• Run 2 (20:15): the mock wall, indoors.
• Run 3 (20:55): a real house, outdoors at night.

All three were recorded on an iPhone 14 without LiDAR. Each report lists issues with stable IDs (A41-1-…, A41-2015-…, A41-2055-…) and maps them to the existing issues #24–#45. The README indexes the three runs.

Why

These are the source for the next round of GitHub issues. Having them in the repo lets the three be compared and deduplicated from any machine before filing.

Run 1's headline findings:

• P0: the ground came from a table. On a phone without LiDAR, the app took the ground height from an ARKit plane labelled table (groundBelow picks the highest plane below the meter). Every height came out about 3 ft 2 in too high.
• P0: the AR view drew nothing. "See it on your wall" showed no battery, cable or tint. This is a regression from AR view draws the battery over people and objects in front of it (R-4) #41's move to RealityKit.
• Looks fixed: "Can't get there" during the walk can put the wall's end at the meter, and the rest of the walk is dropped #24, Marking prompt says "Tap the ac unit" #25 and The "Is this the end of the wall?" prompt never appeared on a 19 ft walk #29.

Verification

• For each run, the packet validator passes and the local solver re-run reproduces the phone's result.
• The narration was transcribed with faster-whisper, and the frames were reviewed at 1 fps plus full-resolution crops at the key moments.

Risk

The repo is public. These are commit-safe copies: meter numbers are removed, and no photos, frames or videos are included. Run 3's media shows a real address and stays in the gitignored private/ folder.

Checklist

• The scope is focused; related changes belong together
• I explained what changed and why
• N/A before/after images and video: docs only, no app changes

🤖 Generated with Claude Code

[chatgpt-codex-connector]

You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard.
To continue using code reviews, you can upgrade your account or add credits to your account and enable them for code reviews in your settings.

[coderabbitai]

Important

• 🔍 Trigger review

This repository does not receive automatic reviews because it has fewer than 10 stars.

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Advanced

Run ID: b1f811f2-3cf8-44fb-9dee-4f5afdb8bb33

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[codemetrics-ai]

Note

Review skipped

Codemetrics is installed on this repository, but it hasn't been added to your team yet.

To enable AI code reviews for this repo, go to your dashboard and add house-scanning-master to your team's repositories.

[AidenJohnston]

Build 4.1 fixes: merge order (final, 2026-09-27)

This plan is superseded. Every build 4.1 fix is now integrated on beta (a4479c0), so no merge order remains to follow for beta.

• Build 5.1 (TestFlight, beta 5bc81fb) carries ios: deflate the text entries in the shared scan zip #87, ios: "leaves out N ft you walked" only on walk tasks, and on the end question #99, ios: draw the AR result with the Canvas until RealityKit is seen drawing it #100, ios: card replies name the task, stay under gate coaching, and lock after a new card #102, ios: gate coaching keeps the task card, holds through dark flicker, and says "Turn more slowly" for turns #104, ios: ground needs a plausible meter height and never rises; estimated meter taps are checked and re-fitted to the detected wall #108, ios: gap-loop follow-ups, upload ticks, the nearer unseen end and request stretches (#39, #31, #83, #75) #109, ios: log meter-anchor corrections so marks can be checked against drift #119 and ios: hold a walk aim until it is met or stalls, and ask for the meter ground only near the meter #120.
• Merged onto beta since 5.1: ios: aim ring fills as its stretch is captured, with a legend and an unambiguous arrow #117, Review: soft gate, window Not sure, and an assumed 3 ft AC footprint #131, ios: review follow-ups for the build 4.1 fixes on beta #132, ios: confirm before a repeated Can't get there ends the scan; no spot for a wall never walked #135, fix(ios): end a side at the phone; confirm and back out of the next wall (#71, #70) #136, ios: review follow-ups for #132 (AR double draw, result chevron, aim legend contrast) #137, fix(ios): bound an end at the phone by the evidence; settle tracking; say 'you saw' #147 and ios: make beta green: aim legend layout and review-flow test waits #155. fix(ios): end a side at the phone; confirm and back out of the next wall (#71, #70) #136 went in before ios: confirm before a repeated Can't get there ends the scan; no spot for a wall never walked #135, as the plan required.
• Still open for t3/ios-mvf: ios: deflate the text entries in the shared scan zip #87, ios: "leaves out N ft you walked" only on walk tasks, and on the end question #99, ios: draw the AR result with the Canvas until RealityKit is seen drawing it #100, ios: card replies name the task, stay under gate coaching, and lock after a new card #102, ios: gate coaching keeps the task card, holds through dark flicker, and says "Turn more slowly" for turns #104, ios: ground needs a plausible meter height and never rises; estimated meter taps are checked and re-fitted to the detected wall #108, ios: gap-loop follow-ups, upload ticks, the nearer unseen end and request stretches (#39, #31, #83, #75) #109, ios: log meter-anchor corrections so marks can be checked against drift #119 and ios: hold a walk aim until it is met or stalls, and ask for the meter ground only near the meter #120 are already on beta. Each needs the caretaker's review-clear before it merges into t3/ios-mvf.
• Server: server: stop requests at the wall's ends, keep placeholder distances out of them, name things without ids #122 targets t3/server and has green CI. It waits on the server deploy.
• Beta CI: iOS run 36313461235 on a4479c0. Before it, the only known failure was the practice-meter UI test, which also leaves its switch on and so fails the onboarding accessibility audit.
• Decisions: proposals and the team's decisions are recorded on AR view draws the battery over people and objects in front of it (R-4) #41, "Looks complete" sends with the window's open/shut answer and the ground type unanswered #65, Corner flow: no "the wall just ends" answer, the refusals loop, and a surface behind the end post was accepted as the next wall #70, Wall ends still land in the wrong place: at the meter on a short side, or 13 ft from where the phone went #71, An AC is one tap, modelled as a 1 ft square, so the battery was placed under the AC stand-in #72, An unwalked, tapped wall still gets a spot (here inside the house) with wall backing and cable route passing #76, 0.16 ft/ft drift, added per feature, makes any spot more than a few feet from the meter unresolvable #79 and No way to finish a scan early: stopping took 7 rapid "Can't get there" taps, each silently setting ends #82.

🤖 Generated with Claude Code

[SamGu-NRX]

@greptileai

[greptile-apps]

Confidence Score: 4/5

[Low risk] Adds field test analysis documentation for a beta build.

The documentation appears safe to merge, with two non-blocking corrections that would make it more reliable for issue triage.

Findings

1.  Exported height direction is reversed ▶
2.  Cross-run references point to private files ▶

Fix with agent prompt

### Issue 1
experiments/device-field-test/build-4.1/run1-2000.md:33-34
The headline says exported heights are about 3 ft 2 in too high, but the report’s measurements show they are too low: the meter is exported at 1 ft 8 in versus about 4 ft 10 in in reality, and the window bottom at 0 ft 2 in versus about 3 ft 4 in. The selected ground plane is too high; the objects’ heights above that plane are too low. The issue list repeats the error, which could send the next round of fixes in the wrong direction.

### Issue 2
experiments/device-field-test/build-4.1/run2-2015.md:7-8
This reference sends readers to `private/device-test-4.1-2000/REPORT.md`, but the public report containing those IDs is `run1-2000.md`. Because `private/` is gitignored, readers of the committed reports cannot follow that path to compare findings. Run 3 likewise names private paths for both earlier reports; link to their public reports instead.

---

For each issue above, determine whether it is valid and should be fixed. If so, fix it directly.

Summary

Adds an index and three commit-safe reports for TestFlight build 4.1 field tests. The reports catalogue capture, guidance, and result issues across two indoor mock-wall runs and one unfinished outdoor run. The first report reverses the direction of its exported-height error, and the later reports point cross-run readers to private rather than public report paths.

_{Reviews (1) · Last reviewed commit: "docs: add build 4.1 field-test runs 2 an..."}