Base x AITX
The weekend, replayed
From Fri 25 Sep, 8:00 PM to Sun 27 Sep, 12:00 PM, Austin time.
593 of Sam’s commits, as he wrote them202 by teammates, counted
- One of Sam’s commits
- Teammates’ commits by the hour, 202 in all, unnamed
- Night in Austin, sunset to sunrise
BaseScanning’s weekend on a 48-hour dial, Fri 12:00 PM to Sun 12:00 PM, Austin time. Each commit is a dot in the ten minutes it was made, stacked outward. Sam made 593 commits, the first at Fri 8:43 PM and the last at Sun 10:00 AM. The busiest hour was Sat 3 AM, with 68. The longest gap between two of his commits was 1 h 54 m, from Sun 6:05 AM to Sun 7:59 AM. Teammates made 202 commits more, counted by the hour and not named. Night in Austin, sunset to sunrise, is shaded. It ended in 1st, Most Commercializable, Base x AITX.
Hour by hour
Fri 8 PM2 commits
docs: add project overview and first capture experimentdocs: match the README to the team plan on main
+ 1 commit by teammates
Fri 9 PM7 commits
Update README.mdMerge pull request #1 from SamGu-NRX/codex/project-readmechore: add iOS, server and web toolchains with CIMerge pull request #2 from SamGu-NRX/t3/setup-toolchainsMerge pull request #3 from SamGu-NRX/codex/live-survey-hldfix(site): pin landing submodule to the implemented pageAdd a scoring harness for capture pipelines against a tape survey
+ 1 commit by teammates
Fri 10 PM8 commits
Reject surveys with uneven checks and runs with conflicting capture timesScore the route check against its review bandexperiments: add measure-lab, a no-LiDAR measurement rigGive the outcome categories one meaning for every checkci: correct the Swift version note in measure-lab.ymlmeasure-lab: keep warnings, signs and keyframes that review found lostmeasure-lab: add a placeholder app iconci: add manual TestFlight upload workflow
Fri 11 PM3 commits
ci: confine the TestFlight key and require approval per runci: add the run attempt to the TestFlight build numberci: scope the TestFlight key to each xcodebuild step and guard build inputs
+ 1 commit by teammates
Sat 12 AM4 commits
fix: validate scoring policies, evidence, and numeric boundsMake passing strict and flag decisions made without their measurementImport Measure Lab sessions as results filesAlign Measure Lab boundary example with Lane C
Sat 1 AM1 commit
Merge pull request #5 from SamGu-NRX/codex/landing-integration
Sat 2 AM8 commits
fix: validate imported quantities and flag blind route decisionsfix: reject invalid measure mappings and duplicate endpointsios: add HouseScanKit package, UI test target and the engine/UI contractserver: publish scene and result JSON schemas (contracts C1, C2)verification: add Simulator runner, STATE probe and UX checklistios: split entry points for the engine and UI lanesfeat(evals): convert a real ADVIO outdoor walk into a Measure Lab replay sessionverification: time states by the log clock; fix probe argument
Sat 3 AM68 commits
server: result schema gains rule_requires_review and clarifies auto_approvedocs: trim AGENTS.md to what agents need and fix stale decisionsios: align the contract with the server's scene and result schemasverification: check replay sessions against C3 and ground truthcontract: two-tap driveway and fence, check needsPerson, result summary and policy flagcontract: result schema fields and two-tap driveway and fenceios: contract additions from the verify lane's UX checklistcontract: skipped cells, cannotAccessArea, check measurements, unseen side, close-up attempts
Show 60 more commitsHide 60 commits
contract: skipped cells, cannotAccessArea, check measurements, unseen side, close-up attemptsserver: placement solver, rules and scene ingestfeat(evals): measure ARKit outdoor drift on ADVIO against three referencesverification: add end-to-end server checkci: trim TestFlight workflow comments and Distribution docsdocs: treat exit code 137 as out of memoryverification: probe checks the Simulator can read a host-path replay; document checksserver: encode the 14 public golden tests; route and evidence fixesClean up the scoring harness without changing behaviormeasure-lab: remove dead code, duplicated checks and process notesverification: solver shim, per-use worktrees, margin rule for other unsure causesverification: record load average with e2e latencyserver: SVG site plan of a placement result (docs/01 D2)HouseScanKit: camera and wall geometry, coverage map, auto-capture, close-up gate, guidance, gap plannerverification: add the scoreboard; e2e reports app exportsverification: regenerate scoreboardHouseScanKit: scene.json export, store-only zip writer, strict placement result decodingserver: defer unseen-stretch geometry to the reported spot; READMEui: every screen of the scan flow, and a scripted demo engineverification: independent placement cases from the public goldensserver: FastAPI layer, property tests, mirror-symmetric sweepverification: count contract problems apart from case expectationsui: frosted fog, card replies, wider demo cameraverification: accessibility audit harness; probe gains a negative controlfeat(evals): score ETH3D reconstructions by point-pair length error against the laser scansci(evals): run the evals unit tests, lint and format on changes to experiments/evalsRuntime: scan engine, live ARKit and replay sources, autopilot, upload, full-flow UI testui: aimed demo cameras, rough walk-to distances, centered status screensverification: fix cases that missed public rules; accept a named raw-body endpointHouseScanKit tests for geometry, coverage, auto-capture, guidance and gap planning; CI runs themserver: publish the scene request body in /openapi.jsonverification: independent accessibility audit per screenverification: pick the JSON-answering upload endpoint; variants keep bundle imagesverification: scope app evidence to the app branch's runs; regenerate scoreboardReplay: load and plan off the main actor; screen held-back windows on precomputed sightingsLive: send the pose every second frame so overlays follow the camera at 30 fpsios/README: describe the engine, launch arguments and how to run the testsResult: place the spot and cable route relative to the meter in the result's own axesui: pass the accessibility audit's type, clipping and contrast checks on text screensios: app icon, accent and named colorsSynthetic replay fixture for CI, and replay decoding testsverification: goldens 07 and 09; cases can rule out one outcome and point at scenes outside gitRuntime: drop an unused server-missing queue and a parameter nobody passedGap planner maps the server's missing evidence; past-end requests clear that endverification: real-geometry scene from ETH3D facadefeat(evals): model runners, ETH3D results, and fixes from an independent reviewEngine: a closed gap advances only if the same request is still showing; log the skipped taskverification: regenerate scoreboardGuidancePlanner: say that a completed task may switch before the dwellserver: fix the independent review's findingsui: keep camera chrome inside the screen at the largest text sizesdocs(evals): report drift and wall-reconstruction results with plain answersui: warmer 3D wall, tidy demo engineEngine: fixes from an independent review of the runtimeserver: route and measurement semantics found by the verification harnessserver: checks with a review band publish review_threshold_ftui: -uiDemoNoFeed for auditing chrome without the photo; simpler reply buttonfeat(evals): score which photos to keep by capture distance and viewing angledocs(evals): point the answers at all three result sectionsui: fog still fades with Reduce Motion, without the drift
Sat 4 AM34 commits
verification: C5 check reads the review line; margin must be within error of a lineui: shared overlay, ripple and press style in their own filesui: one column of marking chips at accessibility text sizesui: no haptic per walk photo, only for deliberate capturesserver: second review roundserver: a route clearly between the reach lines is rule_requires_reviewui: fix review findings in motion, fog and geometryserver: compute ground-patch buffers once per solve
Show 26 more commitsHide 26 commits
ui: close-up flash lights the camera under the chromeUI test: HOUSESCAN_AUDIT_REPORT_ONLY lists every audit issue in one local runAutopilot gate: the UI test releases each screen after screenshotting and auditing itMerge the engine lane: HouseScanKit, runtime, autopilot and full-flow UI testserver: AR default errors grow with distance walked; sweep runs stay adjacentHouseScanKit tests: split an expression Swift 6.2 can't type-check in timeserver: cleanup pass; README and schema text say what a caller needsui: solid Ink chrome over the cameraverification: runner waits only for compiling builds; reports keep every app log lineverification: start the product description of the capture appverification: product description pilot, first Simulator pass, triage; scoreboard evidenceverification: make sim-app and make e2e; reports record peak memoryMerge the UI lane: every screen, app icon, design system and demo modeRuntime: upload to the server's real route, POST /v1/placementsHouseScanKit: accept result fields the server adds within schema 1.0ui: checklist fixes for tracking loss, relocalizing, window question and panel noteUI tests: audit every screen state from the demo engine, and walk the flow through the buttonsMerge the UI lane's checklist fixesUI test: audit the replay flow's settled screens, failing on issues that persistReplayPlanning: put the assumed meter where the walk covers the wallexperiments(meter-closeup): scaffold Vision reader and Commons crawlerfeat(evals): test AR pose priors; fix the fusion comparison and failure accountingui: the AR view names an unsettled spot as the one an installer will checkMerge the ADVIO gap-loop fix: place the assumed meter where the walk faces the wallEngine: walk coaching says "Slow down" and only after the problem lastsrefactor(evals): tidy phase-2 code without changing behavior
Sat 5 AM34 commits
ios/README: document the demo mode and the screen-state audit testserver: deploy to Vercel as its own projectdocs: match rule sources across docs, restore lane ownership, drop the last open Swift/Expo lineui: thinner fog, so the wall shows through where the phone hasn't looked yetcontract: ask what is at a marked wall endverification: product description and UX review at t3/ios-mvf 0876e03; cleanupfeat(evals): score a Measure Lab field session's learned-depth rows against the tapeengine: ask what is at a marked wall end
Show 26 more commitsHide 26 commits
ui: ask what is at a marked wall end, and label the sample result everywhereMerge the engine side of the wall-end questionguidance: aim a walk ahead of the homeowner, not at the first holeexperiments(meter-closeup): label 80 Commons meter photos and measure clean read ratescoverage: a cell is covered only when every row is seen twiceexport: a fence's depth is its nearer tap, not the meanui: the upload steps don't claim to send or check anything for a sample resultengine: measure the ground from planes as they arriveengine: a lost world frame resets only a scan still being capturedengine: a past_end request settles when its end is marked againengine: fix the review's minor findingsserver: fix the second review's findings; serve the schemasserver: cap validation messages instead of echoing the inputserver: cut only the quoted value in validation messagesexperiments(meter-closeup): degradation sweeps, retake thresholds and write-upfix(evals): stream depth maps to disk and cap GPU memory in the model runnerverification: recheck at t3/ios-mvf 525ea40; fix a race in the audit harnessverification: scoreboard with the slow probes runexperiments(meter-closeup): pre-register number-finding criteria; add barcode detection to meterocrMerge fixes from the fresh review: coverage rows, ground height, relocalization, past-end, fence depth, guidance arrowfeat(evals): measure a current iPhone's ARKit scale on MARViN's outdoor walksexperiments(meter-closeup): number-finding rules, frozen on the odd-numbered photosexperiments(meter-closeup): fix a wrong expectation in the tokenizer testfix(evals): make the field kit's AR scale and tape rows soundfeat(evals): report ADVIO position error with the scale share split outRuntime: drop the stand-in answerWallEnd now that every conformer has one
Sat 6 AM14 commits
feat(evals): report position error for the current iPhone on MARViNdocs(evals): report position error for ADVIO and MARViN against the server's allowanceEngine: hold walk instructions for 3 s of screen time, not recording timeexperiments(meter-closeup): barcodes, number finding, retake port spec and field testverification: regenerate scoreboardfix(evals): give the pose-prior verdict intervals and correct its inputsrefactor(evals): name the interval formatter instead of suppressing a lint ruleverification: recheck at t3/ios-mvf beede15
Show 6 more commitsHide 6 commits
server: fixes from CodeRabbit's review of #11MARViN: walk-to-walk spread of ARKit's scale within each siteUI tests and strip: tolerate transient overlays, keep the strip's marks inside and scalableexperiments(meter-closeup): phone-side checks, keyed digests, strict labels, review fixesverification: derive coverage radii from rules.yaml, add ordering, completeness and hostile-input checksCoverage eval: pass criteria and the Swift driver, before any run
Sat 7 AM42 commits
Coverage eval: the app's CoverageMap against what ETH3D's photos sawCoverage eval: a band with nothing claimed reads untested, not passedexperiments(panel-label): pre-register questions; crawler and Vision readercontract: final upload failures, back to review, and the meter-number pickerCoverage options: pre-registered bar and settings, before any runCompile against contract 1cc297f with marked placeholders for the lanes to replaceverification: widen each clearance's reach by the battery's error, check requests per band, bound hostile inputs by memoryRuntime: the interface between the close-up flow and the meter-number reader
Show 34 more commitsHide 34 commits
Coverage: rebuild from the kept cameras when the wall frame changesUpload scene.json only, classify failures, back to review, export survives tap geometryResult decoder: read reason codes, policy sources and crossing effects as stringsCoverage options: angle diversity, view spacing and a depth test on electro's wallDefault the placement server from Info.plistSchemas: re-vendor from origin/t3/server at 739fb6fSend every capture failure to the failure screenKeep only the current scan's folder on the phoneClose-up gate: fail a try for every 4 s without a photo (B-04)Coverage: say plainly that occlusion is not modelledOnboarding: say that only the wall's measurements are sent and the photos stay on the phoneUnsupported screen: plain words for a stopped camera and an unreadable recordingverification: recheck at t3/ios-mvf a39d0a5 and t3/server 739fb6fverification: read ru_maxrss in kilobytes on LinuxRefused uploads go back to review; the close-up asks which number is on the meterMeter number: read the close-up, let the homeowner pick, never auto-fillField kit: one-page sheet, survey and map templates, and a dry run of make fieldField dry run: cameras 6 m out so every synthetic tap is in frame; all rows score 9 of 9Merge round-2 engine lane E1: JSON upload, failure handling, default server, failures on screen, scan cleanup, meter-number flowMeter reader: port the close-up eval's ranking and retake checks, read with Visionverification: friction audit of the homeowner's path at t3/ios-mvf a39d0a5experiments(panel-label): too few openly licensed panel-label photos; record the searchexperiments(panel-label): correct the search counts in the READMECapture: encode JPEGs off the AR queue, stream the bundle zip, no infinite-plane meter tapsMerge round-2 lane M: the meter-number reader ported from the close-up evalMerge round-2 engine lane E2: coverage rebuild, decoder, schemas, B-04, capture queue, meter raycastRuntime: record which plane the meter tap hit, keep results on screen after a camera failureMerge round-2 UI lane: failure ways on, disclosure, meter-number picker, plain failure screensRuntime: log camera-session failuresHouseScanKit: decide which upload failures are worth retrying, with isolated testsserver: fixes from S4's end-to-end run at 739fb6fserver: at radius 0, unseen ground under the footprint is not clearReview: "Add something" opens the camera to mark it, then returns to the list (B-09)Accessibility audit (S4 at a39d0a5): darker primary fill, no low-contrast or clipped transition frames
Sat 8 AM20 commits
verification: recheck t3/server 779f0c3, hosted and 0bad4de; stale rows cite the run that measured the SHAserver: a capture contract a first upload can meetserver: a stretch that ends at the meter reads "to the meter"HouseScanKit: split two expressions Swift 6.2 can't type-check in timeexperiments(meter-closeup): reflow two module docstringsexperiments(panel-label): list the photos to take on field dayMerge round-2 UI follow-up: add from the review (B-09), audit transition fixes, button contrastcontract: tilt-up step, overhead question, and ground-depth, walk-out and overhead requests
Show 12 more commitsHide 12 commits
HouseScanKit: re-vendor the server contract from origin/t3/server e0ee8d3Coverage: sample ground depth out to 15 ft and export it per spanWalk: tilt-up step, overhead question, and copy for the new requestsDemo and screen states: tilt-up, overhead question, new request kindsAutopilot and fixture: tilt up at the end of the walk, answer "open sky"Copy: say what walking the walk-out line showsCoverage: export the walked path as facing clearance; walk 2 m outCoverage: overhead reach of a tilt-up view, exported with the height seenGap loop: honour the server's ground-depth, facing and overhead requestsTest: supplying exactly what each server request names settles itGap loop: an overhead request closes on a new recorded viewTest: too many coverage spans are joined conservatively to fit the schema
+ 1 commit by teammates
Sat 9 AM12 commits
Merge round-3 lane G: tilt-up step, overhead question, request copy, fixture tilt-up framesMerge round-3 lane P: ground depth, walked-path facing, overhead band, server requests honoureddocs: add the 2026-09-26 morning briefingTest: the synthetic fixture has 41 frames since its tilt-up framesOverhead: the tilt-up step records into the coverage map, shim removedGap loop: an overhead request asks what is above once a view covers itGap loop: a server cell request closes only when its whole span is exportedHouseScanKit: a wall chain that follows cornersTest: the gap overhead question state was pasted into the walk-out entryGap loop: a request's walking line runs to the far end of its spanApp: follow a corner during the walkdocs: rebuild the morning briefing around pictures
Sat 10 AM35 commits
Merge corner walking: the scan follows a corner as the next wall in the chaincontract: the shareable scan bundle; HouseScanKit: a hook to break the walked pathHouseScanKit: keep @discardableResult on recordOverheadMerge pull request #19 from SamGu-NRX/t3/briefingEngine: refuse stale keyframe and close-up work before storage startsREADME: run on your iPhoneUI: Share scan on the result, upload-failed and upload-rejected screensHouseScanKit: walked path joins only poses kept close in time with no break between
Show 27 more commitsHide 27 commits
Engine: credit coverage only for a stored, decodable photo judged on its own frameGapPlanner: meet a reach against the server's out_ft as sentEngine: log the first real session's capture decisions; break the walked path on pausesGapPlanner: meet a reach request over its whole span as exportedEngine: ask for a collapsed driveway or fence to be marked again instead of dropping itEngine: offer the packaged scan bundle for "Share scan"HouseScanKit: sample ground to 17 ft; a deeper request is beyond captureEngine: place the battery spot by its span_ft on the wall chainCoverageMap: report ground past a limit end where both sides were seenserver: demo decisions, a key-guarded private mode, and example callsAutopilot: wait for the stored overhead view before reporting itMerge round-4 lane bMerge round-4 lane aMerge round-4 lane cRuntime: pass frame time to coverage and mark limit ends in the map; UI test: no element screenshotscontract: hidden cells, depth availability and a see-behind step for the LiDAR pathios: restore the XcodeGen-generated project filerecon: reconstruction worker scaffold and pre-registered acceptance criteriaci: let the TestFlight build use HouseScanKit, the app's in-repo capture packageFixture: synthetic LiDAR replay with a bin in front of the wallbeta: merge t3/ios-mvf for the TestFlight buildbeta: merge t3/measure-lab for the TestFlight buildbeta: merge t3/testflight for the TestFlight buildUI: hidden cells, the see-behind step, depth-checked map and the follow-up viewUI: say where to look past an obstruction in rough feet, not to the inchMerge pull request #8 from SamGu-NRX/t3/testflightrecon: working worker, ETH3D acceptance, tests, CI, and the handoff to the server team
Sat 11 AM22 commits
beta: merge main (with #8) for the TestFlight buildInfo.plist: declare no non-exempt encryption for TestFlightRuntime: LiDAR depth and mesh capture, depth-aware coverage, see-behind step, automatic server gap loopUI test: full flow from the LiDAR replayMerge remote-tracking branch 'origin/main' into t3/docs-cleanupUI test: audit a control cut off by the screen's bottom edge once scrolled into viewHouseScanKit: encoders for depth/*.u16, depth/*.conf.u8, depth/index.json and mesh.plyHouseScanKit: depth-aware coverage with hidden cells, and a see-behind task
Show 14 more commitsHide 14 commits
HouseScanKit: facing gap and headroom from the LiDAR mesh, exported as facing and overheadsMerge round-5 lane l1Merge round-5 lane l3Merge round-5 lane l2Merge round-5 lane l4packet: capture packet 1.0 spec, schema, validator and sample buildersdocs: rebuild onboarding around the client and server team splitdocs: align the briefing's decision 2 with 3D modeling staying centralProject: keep the UI test fixtures out of the test bundleRuntime: run the LiDAR path on HouseScanKit's API and delete the shimRuntime: hold "One more view to finish" on the upload screen before a follow-upHouseScanKit tests: split a slow #expect in CornerTestsUI tests: check hidden cells on the LiDAR replay for realiOS README: autopilot's server requests, replay depth, and the scan bundle's files
Sat 12 PM19 commits
server: a wall's source sets its default errorpacket: 1.1 — depth frames, estimated depth with sigma, top-level planes, mesh classification flagpacket: mark the fixture's raw sensor files binaryHouseScanKit: meter frame, packet poses and trajectory quaternionsHouseScanKit tests: vendor the capture packet schema and validate allOfHouseScanKit: write capture packet 1.0 foldersverification: final app run on the hosted API at t3/ios-mvf 194f2ebRuntime: log every guidance request the homeowner was shown, and how it ended
Show 11 more commitsHide 11 commits
Runtime: record the capture packet's streams and per-photo metadataRuntime: Share scan exports a capture packet 1.0Merge packet recording and export (engine)Merge the packet writer (package)docs: restore day-1 reasoning in the onboarding and keep eli5 as a recordserver: coverage gaps under 1/8 in are float noise, not unseenMap3D eval: method and pre-registered bar, before any runPacket merge: delete the shim and the unused round-5 bundle encoders; split a slow modifier chainMap3D eval: the app's 3D map against laser visibility on ETH3D electroFix scoring of absent-feature failures and report guidanceGate new-session frames on reset and tracking cycle
Sat 1 PM11 commits
docs: keep only the current plan and the facts that still holdverification: point the facade scene's rule sources at docs/04server: repoint references to docs/01-03 before #14 deletes themEvals: fix CodeRabbit's review findings and the dangling docs referencesPoint Measure Lab comments to retained conventionsHouseScanKit: point comments at docs/00 and docs/04 instead of the deleted docs/01-02ui: "Mark something" steps aside while the walk asks to look past an obstructionUI tests: no live element reads in the audit helper; wait for the AR button to take tapsdocs: add the server team's plan and tighten sentences for agentsbeta: merge t3/measure-lab (43fda59)beta: merge t3/ios-mvf (eb5f1c0) for the TestFlight build
+ 2 commits by teammates
Sat 2 PM18 commits
Vendor the server contract from t3/server 737bf75Add the ground question to the scan contractserver: wall coverage says how high up the face was seenserver: smoke.py keeps the private key on https and off redirectsui: ask what the ground along the wall is on the feature reviewVendor the server contract from t3/server 930e8e5Merge the ground question UIWait out a system banner before the full-flow re-audit
Show 10 more commitsHide 10 commits
docs: ship shared writing, planning and review skills for all agentsExport wall sources and ground patches; close-up into coverage; cable round cornersMerge wall sources, the ground patch, close-up coverage and the cable routePacket: write capture packet 1.1Merge the capture packet 1.1 writerExpand project description in READMEdocs: keep the latest morning briefing without its correction notesdocs: merge the morning briefing into how-it-works as one pageReport how high up the wall each span was seenMerge the wall height seen per span
+ 2 commits by teammates
Sat 3 PM9 commits
docs: make how-it-works concise and skimmabledocs: mark the upload flow as PR #10 and planned work, and caption every scan stepdocs: add the first phone run and correct rule and detector factsMerge pull request #14 from SamGu-NRX/t3/docs-cleanupClose paths that export coverage, patches or walls nobody sawMerge the coverage overclaim fixes from reviewfix: keep a finished scan through lost tracking; fix PR #10 review findingsMerge the runtime fixes from reviewCI: fetch the server and packet branches for the drift tests
+ 1 commit by teammates
Sat 4 PM6 commits
Close the remaining unseen-claim paths from the second reviewMerge the last coverage overclaim fixes from reviewdocs: set the minimal experiment format and cut remaining overstatementsdocs: say the first phone run exists in how-it-worksdocs: separate the headroom ray offset from a height, and re-point evidence before deleting an experimentMerge pull request #33 from SamGu-NRX/t3/docs-minimal
+ 13 commits by teammates
Sat 5 PM6 commits
fix: end the wall where the phone is, and ask for the ground by the meter firstMerge the field-run fix: wall ends where the phone isserver: an existing battery keeps a new one 3 ft away (issue #27)Vendor the scene schema from t3/server 074bf1dMerge pull request #34 from SamGu-NRX/t3/run1-quick-fixesKeep wall ends near the phone, and objects on written walls
+ 19 commits by teammates
Sat 6 PM9 commits
server: fixes from the real-phone run 2 (#44, #45, #42)Ask the walk for the wall to 4.5 ft; report and request above itMerge the walking height: the walk asks for the wall to 4.5 ftVendor the result schema from t3/server 9e42970server: fixes from the caretaker's review of 9e42970Merge #50: name the checks a person must judge in the unseen summaryexperiments/autodetect: meter brand list and scorer, committed before scoringautodetect: pass criteria, fixed protocol and eval sets before any scored runexperiments: pre-register signals-lab questions and pass criteria
+ 17 commits by teammates
Sat 7 PM10 commits
CI: fail when a vendored schema can't be comparedMerge t3/ios-mvf (#53 and later) into the CI drift fixMerge remote-tracking branch 'origin/main' into t3/reconMerge remote-tracking branch 'origin/main' into t3/evalsReplace field-run values in a test and the demo with made-up onesrecon: wall needs a surface to count as seen; the wall line carries its own error; content-keyed MoGe cacheexperiments: run the first signals-lab roundfix(scoring): protect imports and align synthetic session evidenceMerge main into scoring experimentMerge remote-tracking branch 'origin/main' into t3/meter-closeup
+ 4 commits by teammates
Sat 8 PM3 commits
packet: point to the server's capture packet 0.4, keep our 0.5 proposalsautodetect: Vision rectangles and OWLv2 scored, laser ground truth for door extentspacket: list the validator hazards #22's review found, for the server's validator
Sat 9 PM10 commits
CI: fetch the contract branches unless the compared files are checked outexperiments: move method detail into METHODS.md so each README is a short front doorDrop the live packet 1.1 schema comparison; t3/packet retired itFix the #10 caretaker's findings on the anchor, ground, source and scan folderspacket: add the signals lab's 0.5 proposals; describe the checks without the server's pathspacket: fix the proposal list's formattingAdd the live dots prototype, a macOS replay of the holographic dot viewdocs: add the app review (product description, UX review, friction audit)recon: rerun ETH3D electro acceptance after the coverage and oracle fixesRead the result's answer, clean spot, nearest failure and view links in the package
+ 2 commits by teammates
Sat 10 PM20 commits
Tune the live dots for a 1.5 to 3 m working distanceLet simulated feature points land on brick texture, and add resultsautodetect: Create ML student scored, 3D extent results, resource recordMap the nearest spot, the rules notice and each check's view into the resultLead the result with the answer and fold the full checks under DetailsShow the replay and autopilot badge as a quiet pillFill the launch screen with the Canvas colourSwap the upload headline at once instead of crossfading it
Show 12 more commitsHide 12 commits
Merge the engine fixes from the #10 review: anchor turn, ground plane, restart, failure, cleanupTrim the live dots README to the experiment format's lengthverification: fix the review's nine findings; move the app reviews to #61Result model follows corners; Reduce Motion stops moving things; audit keys by identifierMerge the UI fixes from the #10 review: corners in the result model, Reduce Motion, audit keyverification: count a past-end request as covering what lies beyond the endevals: fix the caretaker review findings and split the READMEexperiments(meter-closeup): front-door README; fix sweep resume, leak scanner and OCR inputrecon: carry the evals corrections into HANDOFFAdd a scheme switch to compare three looks on the live dots replayverification: cut the README to the experiment format; move what each check asserts into the docstringsexperiments/autodetect: meter brand results, lab result and integration plan
+ 1 commit by teammates
Sat 11 PM33 commits
Warm ember dots only at birthpacket: mark the proposals unapproved, state the studies' limits and the checks' limitsRetry uploads the server answered 408 or 429Merge main (f852207) into t3/ios-mvfverification: require views strictly past the needed height; credit walk-past requests only at the real endexperiments(meter-closeup): exact sweep completeness, fresh-run directory, full CSV leak scanfix(scoring): reject unsupported route evidence and empty review bandsverification: check each sweep start alone; survive malformed answers and shared builds
Show 25 more commitsHide 25 commits
ios: show the meter's brand beside the number candidatesMerge remote-tracking branch 'origin/t3/ios-mvf' into t3/meter-branddocs: correct the app review against its own documents and checklistverification: say that sweep coverage samples starts, not the server's listpacket: tighten three proposals after the Codex review of 3792a8fGive the Details header an explicit button the UI tests can openexperiments(meter-closeup): scrub a short plate number, keep the key off PRs, harden the sweep cacheevals: second-round review fixesbuild: include available experiment suites in root checksfix(scoring): validate session geometry and bound decompressionDrive the result screenshots and UI tests from server answer filesShow which rules answered and keep the nearest spot at the server's sizeUse the system disclosure for Details and keep one filled buttonci(meter-closeup): run the keyed leak check only on mainrecon: second review round: wall error from the server, per-frame LiDAR, visible-view scoringexperiments(meter-closeup): split the held-out set by physical meterverification: facing and headroom need the band past the battery by the wall's errorMeter close-up: out of focus at 6.68, the eval's current thresholdverification: widen facing and headroom by the drifted error, as the server doesAdd fog that dots replace, recognised boxes and guidance statesbeta: assemble current iOS fixes for simulator QAbeta: include onboarding motion permission for testingbeta: include walk guidance regression coverage for testingbeta: exercise refused-view progression with known coverage limitationsUI tests: read audit elements from one snapshot, never live
Sun 12 AM34 commits
Trim the live dots README back toward the experiment format's lengthfix(scoring): protect report inputs and escape Markdown valuesdocs: link the live site and pin the merged landing pageMerge remote-tracking branch 'origin/main' into t3/reconevals: final review round, including two zip safety fixesMerge remote-tracking branch 'origin/main' into t3/evalsrecon: confine manifest file references to the packet; record known defectsios: 44 pt reject target, and the brand lives only beside its number
Show 26 more commitsHide 26 commits
Merge remote-tracking branch 'origin/t3/ios-mvf' into t3/meter-brandShow the sample badge as a quiet pill under Replayexperiments(meter-closeup): human corrections in the leak digests, safe field-test files, end-to-end rankingverification: lay walls out and check declared heights as server 7133283 doesevals: bound a session archive before hashing or opening itMerge t3/ios-mvf (b9851e1) into t3/experience-resultverification: scoreboard and result at server 7133283evals: bound a session zip's central directory before ZipFile reads itrecon: refuse unsafe keyframe ids; record two more known defectsevals: test the archive size cap with a faked 5 GiB sizerecon: keep MoGe-2 cache files inside the cache folder; cite defect source linesMerge the audit's stable element reads (the #24 test flake)docs: draw the tap-list bar at 27 of 34 to match its labelAsk the homeowner to confirm the spot before showing the resultBuild the sample badge as one Text so it passes the Dynamic Type auditList under Settles only the checks a view settlesNever call a spot with no checks a fit or cleanSize sweep zones by the nearest spot when there is no spotMerge commit '0a3a05e200fa42daf002d5fc77687c898935cffa' into codex/beta-qa-candidate-20260927Merge commit '5150321a22f540e669755270dd76caa23957d8a4' into codex/beta-qa-candidate-20260927Cap Retry-After at a day so the wait message can't overflowKeep "Can't get there" on screen through the capture gate's coachingKeep the last completed scans across a relaunch, and stamp each scanverification: a wall view without out_ft reached headroom height, as server 6200098 reads itfeat(tools): add a local viewer that shows a capture reaching the serverfix(tools): update the viewer's file counts when marks land
+ 14 commits by teammates
Sun 1 AM22 commits
Merge commit 'f14947eb9bff5b707660c8d873b72e60ba056631' into codex/beta-qa-next-20260927Merge commit 'b5006fca782aa280caa767e476fd5c62b496a9e2' into codex/beta-qa-next-20260927Merge commit '9c2a5676fff6b31b1eb1e4bde63a5473ae61b5f3' into codex/beta-qa-next-20260927Merge commit 'a4b8b2723736b5f2b29c5605569dcdb76c6d8a69' into codex/beta-qa-next-20260927Merge commit 'f141a6212f3bba8c26695bf3cb5cb40ef0f1cf9d' into codex/beta-qa-next-20260927Merge commit '5c15da8705ef09bef33a71d1d1dca27deddb7480' into codex/beta-qa-next-20260927fix(tools): keep viewer counts and stage marks current when animations cannot runRegenerate the Xcode project for the result answer fixtures
Show 14 more commitsHide 14 commits
verification: claim only demonstrated reach in the ETH3D builder; owe requests for a spot's unseen checksfix(tools): reconnect within 5 s after the API comes backShow the instruction card's reply without a fadeMerge the spot confirmation: the homeowner checks the chosen spot before the result countsMerge #89: read the meter's brand from the close-upMerge keeping the last scan across a relaunch, stamped with its revisionIntegrate result redesign with beta gap-loop fixesverification: scoreboard and result at server 9176125fix(tools): keep the latest run current while the viewer catches up on historyfix(tools): switch runs on a retake and keep polling current without eventsverification: app rows at t3/ios-mvf f14947e; read its bundle line; tolerance for requestsMerge non-LiDAR AR and aim guidance for beta testingfix(tools): keep the illustration consistent and recover previews, health and resultsMerge repaired walk guidance for beta testing
+ 7 commits by teammates
Sun 2 AM13 commits
Merge reply-card transition repair for beta testingMerge accessible follow-up controls repair for beta testingMerge coaching repairs while preserving beta guidance and repliesfix(tools): name the views the server asks for and drop models it no longer offersverification: close three false certifications; pin to server 4a9fa44fix(tools): show requested views before a result exists and scope them to the runUI test: tap only once the element can take the tapMerge retained scans and spot confirmation into beta candidate
Show 5 more commitsHide 5 commits
Merge visible aim completion while retaining beta coaching and retention copyfix(tools): ignore stale retakes, re-read revised results and previews, pace counts per batchfix(tools): let a run first seen in later events replace the run the status namedMerge reviewed beta follow-ups for anchors and requested spansMerge reviewed AC footprint and feature review follow-ups
+ 10 commits by teammates
Sun 3 AM24 commits
Allow the expanded beta UI suite to finishevals: stage model outputs until a run completes; bound image pixelsrecon: refuse a capture with duplicate keyframe idsrecon: test that duplicate keyframe ids are refused in both formatsfix(tools): pin a status-named run before the first live batch, label stale refreshesPreserve [teammate] beta merge history in the combined candidateverification: sample ground cells at their centres as well as their edgesGive the instruction card's reply button a solid fill
Show 16 more commitsHide 16 commits
evals: all-or-nothing publication of model runs; bound session memoryRestore the spot-refusal notice required by the result mergeexperiments(meter-closeup): scan across every separator, keep confirms corrections, no tied barcode promotionios: practice meter, a drawn sample meter for scanning without oneverification: pin to server 86b2b4d; scoreboard and resultMerge commit '5c83ad49efd108874102b75327798ecd4d654f82' into codex/beta-qa-followups-20260927ios: recheck the scan after drawing the practice close-up photoMerge latest aim layout into beta QA follow-upsPreserve latest beta wall-end and corner integrationDraw the prototype's fog and live dots over the walkrefactor(tools): let the capture status alone decide which run the viewer showsIntegrate practice meter with beta permissions and mode badgesKeep hidden and requested tints in the band, and rag the fog's edgesSay in the iOS README where the live fog's rule and dots liveKeep narrow unseen cells fogged, move dots exactly on correctionsMerge commit '9defff1' into t3/live-fog-beta
+ 25 commits by teammates
Sun 4 AM13 commits
Merge reviewed object-tap reach fix into beta candidateverification: a reject needs every start failing; a footprint needs its wall; pin to a726199Merge [teammate] beta UI layout and test navigation repairstest(ios): isolate Practice and restart through result DetailsReplay: an upper walk that sees the wall above 6.5 ft, and a gap the Can't-get-there flow asks forexperiments(meter-closeup): scan the whole repository for leaks, count unread photos in rule precisionTune the fog, the hidden tint and the dots against the prototype videostest(ios): use synthetic object tap fixtures
Show 5 more commitsHide 5 commits
test(ios): use synthetic object tap fixturesevals: publish grouped MapAnything outputs all or nothing; state the pose-noise confoundMerge current beta recovery and walk-exit fixes into QA candidateMerge reviewed result actions and AR outline correctionsIntegrate onboarding moves while retaining Practice accessibility coverage
+ 18 commits by teammates
Sun 5 AM8 commits
Merge live fog and LiDAR dots into the combined beta candidatePreserve beta history and reviewed ground-request stall repairMerge reviewed beta end-position follow-up into demo candidatechore: Add MIT License to the projectMerge reviewed test scrolling correction for combined beta QAPreserve beta TestFlight stamping repair in demo candidateWall map legend wraps instead of overlapping; picker wait allows a stalled SimulatorPreserve beta merge history for the included aim repair
+ 9 commits by teammates
Sun 6 AM1 commit
Keep the Practice label visible at large text sizes
+ 6 commits by teammates
Sun 7 AM1 commit
verification: pin to server 3baa338; accept a pass by the smallest reported margin
+ 8 commits by teammates
Sun 8 AM4 commits
evals: untested zero-claim options, refuse partial multi-view runs, bind the surveyMerge remote-tracking branch 'origin/main' into t3/evalsevals: correct three statements the final review foundverification: match server 3baa338's coverage, from the final Opus review of 255bc84
+ 9 commits by teammates
Sun 9 AM4 commits
verification: three more false certifications from Codex on 255bc84experiments(meter-closeup): scrub notes as the leak check reads them, keep corrections across rebuilds, digest sweep inputs, measure field photos unrecompressedexperiments(meter-closeup): re-sweep under the input digest; every measurement unchangedverification: app rows at #10's freeze c820cf1; band gaps under 0.01 ft are rounding
+ 6 commits by teammates
Sun 10 AM1 commit
fix(tools): stop inferring placeholder processing from stage durations
+ 24 commits by teammates
Sun 11 AM
+ 1 commit by teammates
Sun 27 Sep, 12:00 PMBack to BaseScanning
