Repository navigation
CI: release workflow bumps version (patch/minor/major) in source before releasing - #8
Merged
Merged
Conversation
…before releasing The release build now takes a `bump` choice input (patch = 0.0.x, minor = 0.x.0, major = x.0.0; plus `none`, and an explicit `version` override). The bump job increments the version IN SOURCE — android/app/build.gradle.kts (versionName + versionCode) and project.yml (MARKETING_VERSION + CURRENT_PROJECT_VERSION) — commits it to the dispatched branch (authored ryanbr), and the platform jobs check out that bumped commit, so the built app actually REPORTS the new version. Then it cuts the non-prerelease v<NEW> release. Each platform's build number increments independently. Verified locally: version math (8.2.2 → 8.2.3 / 8.3.0 / 9.0.0) and the sed edits hit only the version lines (the -debug/-staging/-demo suffixes and $(MARKETING_VERSION) refs are untouched).
2 of 3 tasks
ryanbr
pushed a commit
that referenced
this pull request
Jul 8, 2026
A WHOOP MG that tops up short of full every day never trips the near-full anchor (rule 1), so the discharge fit fell back to scanning the SoC buffer's global max — which, across many partial-top-up cycles, can sit several days back. The endpoints-only fit then nets across every undetected intermediate top-up in between, diluting the actual recent drain rate into something far flatter than reality (reported "9% = ~3 days" when the true recent drain was closer to 6h left). Bound the max search to at most the last two charge-step cycles so it stays anchored to how the strap is being used right now. Buffers with 0 or 1 charge-steps search from the start exactly as before, so #8 and #919 are unaffected — verified against all existing fixtures plus a new multi-cycle regression test. Ported the same fix to the Kotlin twin to keep both platforms behaviour-identical. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
TheBoroer
added a commit
to TheBoroer/ryanbr-noop
that referenced
this pull request
Jul 10, 2026
…WIP) One windowed readings list drives the trend chart, the 'N readings' caption, and the new readings table, so the three can never disagree. Source labels reuse TodayView.provenanceDisplayLabel — no new source vocabulary. Swift + Kotlin twins with projection tests.
TheBoroer
added a commit
to TheBoroer/ryanbr-noop
that referenced
this pull request
Jul 10, 2026
9 of 12 tasks
ryanbr
added a commit
that referenced
this pull request
Jul 16, 2026
…description typos (#492) (#496) * Today (Android): unclip the day title + drop the bogus Strain % (#492) Two Android-only visual bugs from #492 (iOS already handles both correctly): - #1: the tappable day-title Column had .clip(RoundedCornerShape(cornerSm)) with indication = null — no ripple to shape, but the rounded corners clipped the title/date text's bottom-left (the 'W' of 'Wednesday'). Drop the clip; the clickable still defines the tap area. iOS uses contentShape(Rectangle()), no clip. - #8: the Strain Key-Metrics tile hardcoded unit = "%". Strain/Effort is a load index (0–21 WHOOP / 0–100 NOOP), not a percentage, so the % was wrong — most visibly on the 0–21 scale ('5.7%'). Recovery/Rest are genuine 0–100 % and keep it. iOS shows the strain axis as an 'of 21'/'of 100' caption with no %. compileFullDebugKotlin: BUILD SUCCESSFUL. Reported by bartmuskala. * Trends (Android): fix two metric-description punctuation typos (#492 follow-up) Spotted in the #492 screenshots (not in the reporter's list): the Charge/Sleep metric-explorer descriptions had a space before the comma, and the Sleep one also used a comma where iOS uses a colon. These are documented to 'match the macOS MetricCatalog entries exactly' (TrendsExploreScreen :190), so this restores that parity: - 'you are , led' -> 'you are, led' - 'sleep was , dura...' -> 'sleep was: dura...' (matches MetricCatalog.swift) String-literal only; no logic.
DX23876
referenced
this pull request
in DX23876/noop
Jul 20, 2026
… the coaching model Reported: the two background-role model fields (chat summaries, card analyses) were plain TextFields — no picker, no validation against the provider's real catalogue, no "Custom…" affordance, no refresh — unlike the main coaching model, which already has a Picker/searchable-sheet/Custom… pattern (modelSelector) driven by coach.availableModels. roleModelField now reuses that exact mechanism per role instead of a bare TextField: - Below the searchable threshold: a menu Picker listing every model the provider actually returned, plus "Same as coaching model" (empty — the role's real default, kept distinct from Custom) and "Custom…" (typed id outside the fetched list, same two-step tag/TextField dance as the main picker). - Above the threshold (OpenRouter's 300+): the same searchable-sheet button, routed through a new enum-driven ModelSearchTarget (chat/summary/ cardAnalysis) rather than a THIRD independently-toggled `.sheet(isPresented:)` — the exact stacked-modifier shape R2 just fixed for the goal dialogs, so this package doesn't reintroduce it three fields over. - A Refresh-models affordance in the Background models card itself, calling the same coach.refreshModels() the coaching model's button already does (one shared per-provider model list — independent-provider-per-role is #8, still deferred) — repeated here for discoverability in this card. Both roles are independently selectable; empty still means "inherit the provider's cheap model" exactly as before, just chosen from a real option instead of an empty text field. 2 new UI strings ("Same as coaching model", the reworded background-models caption), localized de/es/fr. Both targets build; full suite green (1210, 0 failures); i18n audit exit 0 with 44 pure catalog insertions, 0 deletions. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
DX23876
referenced
this pull request
in DX23876/noop
Jul 22, 2026
CI: release workflow bumps version (patch/minor/major) in source before releasing
DX23876
referenced
this pull request
in DX23876/noop
Jul 22, 2026
…br#99) A WHOOP MG that tops up short of full every day never trips the near-full anchor (rule 1), so the discharge fit fell back to scanning the SoC buffer's global max — which, across many partial-top-up cycles, can sit several days back. The endpoints-only fit then nets across every undetected intermediate top-up in between, diluting the actual recent drain rate into something far flatter than reality (reported "9% = ~3 days" when the true recent drain was closer to 6h left). Bound the max search to at most the last two charge-step cycles so it stays anchored to how the strap is being used right now. Buffers with 0 or 1 charge-steps search from the start exactly as before, so #8 and ryanbr#919 are unaffected — verified against all existing fixtures plus a new multi-cycle regression test. Ported the same fix to the Kotlin twin to keep both platforms behaviour-identical. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
DX23876
referenced
this pull request
in DX23876/noop
Jul 22, 2026
…description typos (ryanbr#492) (ryanbr#496) * Today (Android): unclip the day title + drop the bogus Strain % (ryanbr#492) Two Android-only visual bugs from ryanbr#492 (iOS already handles both correctly): - #1: the tappable day-title Column had .clip(RoundedCornerShape(cornerSm)) with indication = null — no ripple to shape, but the rounded corners clipped the title/date text's bottom-left (the 'W' of 'Wednesday'). Drop the clip; the clickable still defines the tap area. iOS uses contentShape(Rectangle()), no clip. - #8: the Strain Key-Metrics tile hardcoded unit = "%". Strain/Effort is a load index (0–21 WHOOP / 0–100 NOOP), not a percentage, so the % was wrong — most visibly on the 0–21 scale ('5.7%'). Recovery/Rest are genuine 0–100 % and keep it. iOS shows the strain axis as an 'of 21'/'of 100' caption with no %. compileFullDebugKotlin: BUILD SUCCESSFUL. Reported by bartmuskala. * Trends (Android): fix two metric-description punctuation typos (ryanbr#492 follow-up) Spotted in the ryanbr#492 screenshots (not in the reporter's list): the Charge/Sleep metric-explorer descriptions had a space before the comma, and the Sleep one also used a comma where iOS uses a colon. These are documented to 'match the macOS MetricCatalog entries exactly' (TrendsExploreScreen :190), so this restores that parity: - 'you are , led' -> 'you are, led' - 'sleep was , dura...' -> 'sleep was: dura...' (matches MetricCatalog.swift) String-literal only; no logic.
DX23876
referenced
this pull request
in DX23876/noop
Jul 22, 2026
… the coaching model Reported: the two background-role model fields (chat summaries, card analyses) were plain TextFields — no picker, no validation against the provider's real catalogue, no "Custom…" affordance, no refresh — unlike the main coaching model, which already has a Picker/searchable-sheet/Custom… pattern (modelSelector) driven by coach.availableModels. roleModelField now reuses that exact mechanism per role instead of a bare TextField: - Below the searchable threshold: a menu Picker listing every model the provider actually returned, plus "Same as coaching model" (empty — the role's real default, kept distinct from Custom) and "Custom…" (typed id outside the fetched list, same two-step tag/TextField dance as the main picker). - Above the threshold (OpenRouter's 300+): the same searchable-sheet button, routed through a new enum-driven ModelSearchTarget (chat/summary/ cardAnalysis) rather than a THIRD independently-toggled `.sheet(isPresented:)` — the exact stacked-modifier shape R2 just fixed for the goal dialogs, so this package doesn't reintroduce it three fields over. - A Refresh-models affordance in the Background models card itself, calling the same coach.refreshModels() the coaching model's button already does (one shared per-provider model list — independent-provider-per-role is #8, still deferred) — repeated here for discoverability in this card. Both roles are independently selectable; empty still means "inherit the provider's cheap model" exactly as before, just chosen from a real option instead of an empty text field. 2 new UI strings ("Same as coaching model", the reworded background-models caption), localized de/es/fr. Both targets build; full suite green (1210, 0 failures); i18n audit exit 0 with 44 pure catalog insertions, 0 deletions. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
brendankowitz
pushed a commit
to brendankowitz/noop
that referenced
this pull request
Aug 9, 2026
The mockup's own note for this screen is "the one-line flat-ink fix, plus a
provenance sentence in the header" — this is that, plus the two group cards the
tree can honestly fill.
- ScreenScaffold now takes `topBackground: liquidFlatInkBackground()`: Data
Sources is an archive/utility page, not a lived moment.
- Header sentence is built only from figures the screen already prints (the WHOOP
card's own days/sleeps counts) and closes with this screen's existing on-device
line, verbatim. NOT the mockup's "four sources feeding today, everything current
except Apple Health, last read yesterday": nothing in the tree records a
per-source last-read time for the file importers, so that sentence would be
invented provenance.
- The live-strap status card becomes GroupCard("Connected") with the mockup's
status dot — the same three-state read the Live screen's pill uses (ryanbr#8), sage
bonded / amber connected-not-streaming / terracotta offline — and the one real
"synced N ago" the app has (`live.lastSyncedAt` + `relativeAgo`, the pairing the
Health screen's sync status already renders). The file importers get no dot:
they have no connection state to report.
- New ink ActionCard "Sync now" reaching the SAME gated entry point Health's
Sync-now button uses (`BLEManager.syncNow()` is a no-op when no strap is
connected or a sync is running), disabled with an honest subtitle when it can't
run.
- GroupCard("Stored on <device>") carries the real days / sleep-sessions counts.
- Every importer card, binding, picker and destructive action is untouched.
Verified: xcodebuild -scheme Strand -destination platform=macOS build succeeds.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01GPcFU4xvYzxovzd4aQ7uGB
simoncad7
pushed a commit
to simoncad7/noop
that referenced
this pull request
Aug 17, 2026
…description typos (ryanbr#492) (ryanbr#496) * Today (Android): unclip the day title + drop the bogus Strain % (ryanbr#492) Two Android-only visual bugs from ryanbr#492 (iOS already handles both correctly): - ryanbr#1: the tappable day-title Column had .clip(RoundedCornerShape(cornerSm)) with indication = null — no ripple to shape, but the rounded corners clipped the title/date text's bottom-left (the 'W' of 'Wednesday'). Drop the clip; the clickable still defines the tap area. iOS uses contentShape(Rectangle()), no clip. - ryanbr#8: the Strain Key-Metrics tile hardcoded unit = "%". Strain/Effort is a load index (0–21 WHOOP / 0–100 NOOP), not a percentage, so the % was wrong — most visibly on the 0–21 scale ('5.7%'). Recovery/Rest are genuine 0–100 % and keep it. iOS shows the strain axis as an 'of 21'/'of 100' caption with no %. compileFullDebugKotlin: BUILD SUCCESSFUL. Reported by bartmuskala. * Trends (Android): fix two metric-description punctuation typos (ryanbr#492 follow-up) Spotted in the ryanbr#492 screenshots (not in the reporter's list): the Charge/Sleep metric-explorer descriptions had a space before the comma, and the Sleep one also used a comma where iOS uses a colon. These are documented to 'match the macOS MetricCatalog entries exactly' (TrendsExploreScreen :190), so this restores that parity: - 'you are , led' -> 'you are, led' - 'sleep was , dura...' -> 'sleep was: dura...' (matches MetricCatalog.swift) String-literal only; no logic.
ryanbr
added a commit
that referenced
this pull request
Aug 27, 2026
Re-review found a fourth didBond reader I had missed. CLAUDE.md warns that didBond is load-bearing well beyond the handshake; I audited the bond watchdog, the #982 detector and the give-up, and stopped there. The keep-alive timer reads it too. keepAliveFire gated on didBond, which a suppressed 5/MG never reaches. That switched the whole timer off for exactly the link the suppression keeps up for hours, taking the liveness watchdog with it - so a stalled stream would have had nothing left to bounce it, and the "stable live HR" this port promises would have quietly depended on a stream nothing was watching. The Android twin gates on `bonded`, the live-HR shortcut, and so keeps reaching its watchdog. Worse, the first attempt at this was dead code: BOTH existing startKeepAlive() call sites are post-bond, so on a suppressed link the timer never starts at all and a corrected gate changes nothing. It now starts at the suppression point too. Safe before HR arrives, because the gate returns false until the stream marks the link established (#8). Deliberately narrower than Android's gate: only a 5/MG is admitted unbonded, and only as far as the watchdog - a second didBond guard closes the door before anything puffin-framed, which an unbonded link cannot land anyway. A 4.0 sitting unbonded is a real fault, not a suppressed link, and must not be kept alive. Verification: the gate is extracted as pure keepAliveMayRun so it is pinned without a radio - 4 tests incl. every DeviceFamily case either side, and the whole pure file re-compiled standalone with swiftc and every rule exercised. doc_comment_lint clean. Still not hardware-validated.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What
Updates the manual "Release build (fork)" workflow to bump the version before releasing.
New
bumpchoice input — patch (0.0.x) / minor (0.x.0) / major (x.0.0) / none, plus an optional explicitversionoverride. The bump job edits the version in source —android/app/build.gradle.kts(versionName+versionCode) andproject.yml(MARKETING_VERSION+CURRENT_PROJECT_VERSION) — commits it to the dispatched branch (authored ryanbr), and the platform jobs build that bumped commit so the app reports the new version. Then it cuts the non-prereleasev<NEW>release.Notes
-debug/-staging/-demosuffixes and$(MARKETING_VERSION)refs untouched).testing-latestsnapshot; still a non-prerelease soreleases/latest(the update check) resolves against it.