Repository navigation
"N checks depend on areas the scan did not see" counts checks that are unsure for other reasons #45
Description
Seen in device field test run 2: TestFlight build 3.1, beta at ad91620, capture code t3/ios-mvf e9e405d, server e106acb. The exported scan was lost before it was copied off the phone, so this comes from the screen recording and the code at ad91620. Times are in the recording. b7f51ba and 49a676e on t3/ios-mvf don't change this.
What happens. The result read: "More views are needed around the best spot, 1 ft 3 in left of the meter: 10 checks
depend on areas the scan did not see." One of the 10, "Clear of the meter's working space", is UNSURE because of
measurement error, not unseen area. More views won't settle it.
Cause. The summary switches to this sentence if any UNSURE check is unobserved, but counts all UNSURE
checks (ids = best.unsure(); solver.py#L1369, solver.py#L1393-L1398).
Possible fix. Count only the checks whose unsure_cause == "unobserved", and add a clause for the rest (for example
"and 1 is too close to call").
Activity
server: fixes from the real-phone run 2 (#44, #45, #42)
SamGu-NRX commentedon Sep 26, 2026
Fixed on
t3/server(#11) in 9e42970, live on both hosted servers. The sentence counts only theunobservedchecks and adds the rest as a count: "More views are needed around the best spot, …: 1 check depends on areas the scan did not see, and 1 needs a person to judge." Test:test_the_summary_counts_unseen_checks_and_names_the_rest. Note that #50 fixes the same lines differently: it names the other checks ("; a person also needs to check …") instead of counting them. The two now conflict. Whoever merges should pick one wording; #50 would need rebasing onto 9e42970.Merge #50: name the checks a person must judge in the unseen summary
AidenJohnston commentedon Sep 27, 2026
Build 4.1 field test (TestFlight 4.1,
beta45017ba, iPhone 14 without LiDAR, three runs on 2026-09-26): looks fixed.The summary now counts only unseen checks and lists the rest separately:
The same summary still ends with the demo-rules notice, and a check reads "objects[1] ac"; both are in #74. Analyses: PR #60 (run 1, run 2, run 3).
Align the capture-packet schema with spec 0.4.1. (#45)