Defense-in-depth against GPU init failures on Snap/Flatpak Linux where
the main process stays alive but the GPU process crashes at init and
the window never renders. Field data in #7270 (two post-v18.2.4
reports) shows this happens on Ubuntu 24.04+/25.10 regardless of GPU
vendor — the driver is core22 Mesa/libgbm drifted from the host Mesa.
See §12–§17 in docs/research/snap-wayland-gpu-fix-research.md.
Mechanism (electron/gpu-startup-guard.ts):
- Content-based crash marker in userData with {ts, electronVersion}.
Written before app.whenReady() on confined Linux; cleared via
IPC.APP_READY after Angular boot — not ready-to-show, which fires
on blank/broken renderers too.
- Previous-crash detection: marker present AND recent (<5 min) AND
matching Electron version. Staleness bound + version gating drop
systemd-SIGKILL-mid-boot and post-upgrade-residue false-negatives.
- Env overrides SP_DISABLE_GPU=1 / SP_ENABLE_GPU=1 work on all
platforms; auto-detection is Linux+Snap/Flatpak-only.
- Non-ENOENT fs errors logged at warn — a swallowed write-fail
previously meant the guard could re-enter the loop with no
diagnostic trail; a swallowed unlink-fail meant a successful boot
could get permanently stuck in crash-recovery.
Fallback flag bundle (start-app.ts):
--disable-gpu
--disable-software-rasterizer
--ozone-platform=x11
The pair matches Chromium's GPU integration tests' "no GPU process"
invariant; DisplayCompositor handles 2D in the browser process
without spawning a GPU child. app.disableHardwareAcceleration()
alone does NOT — verified against electron/electron#17180/#20702.
The extra --ozone-platform=x11 closes the Chromium 140+
browser-side Wayland/libgbm-dlopen gap on Flatpak (redundant with
the existing Snap X11 widening branch; last flag wins).
Novelty: R3 survey of VS Code, Slack, Insomnia, LosslessCut,
Obsidian flatpak, Firefox snap, and Canonical's gpu-2404-wrapper
found no peer Electron-snap implementing an equivalent reactive
crash-detection + auto-fallback. Prevailing patterns are manual
--ozone-platform=x11, proactive env-sniff, or do-nothing.
Review: §14–§15 multi-agent verification on the original PR #7273;
re-reviewed via 7-agent multi-review (6 Claude focus-agents + Codex
CLI) + 5 research agents here (§17). Live-tested with
SP_DISABLE_GPU=1 on a KDE/X11 dev host — window rendered normally
via DisplayCompositor.
Refs: #7270
* fix(issue): prevent crash from orphan issueProviderId (#7135)
The Jira image-headers effect in task-detail-panel subscribed to
selectIssueProviderById without an error handler, so a task with an
issueProviderId pointing at a deleted provider (e.g. after sync
convergence where taskIdsToUnlink didn't cover all local tasks)
propagated the selector throw to Zone.js as a crash dialog. Wrap the
inner selector observable in catchError that logs and falls back to
of(null); the downstream jiraCfg?.isEnabled guard handles the fallback.
Also drop IssueLog.log(issueProviderKey, issueProvider) from the
throwing variant of the selector: providers may carry credentials
(host, token, apiKey) and IssueLog history is exportable.
* fix(focus-mode): sync tray countdown with in-app timer during breaks
Tray title was rebuilt from a cached currentFocusSessionTime that only
refreshed when CURRENT_TASK_UPDATED fired. addTimeSpent is gated on an
active current task, so during focus-mode breaks or task-less focus
sessions the cache froze while the in-app timer kept ticking.
Add the tick action to taskChangeElectron$ so the cache refreshes every
second whenever the focus timer is running.
Fixes#7278
* fix(ci): restore GitHub Actions SHA pins undone by 0e9218bd68
Commit 0e9218bd68 silently reverted PR #7212 (github-actions-minor group
bump) along with its stated sync/client-id work. This restores the 15
workflow files to their pre-revert state.
Actions restored to newer pinned SHAs:
- actions/upload-artifact v7.0.0 -> v7.0.1
- step-security/harden-runner v2.16.1 -> v2.17.0
- softprops/action-gh-release v2.6.1 -> v3.0.0
- signpath/github-action-submit-signing-request v2.0 -> v2.1
- anthropics/claude-code-action v1.0.89 -> v1.0.93
- docker/build-push-action v7.0.0 -> v7.1.0
- easingthemes/ssh-deploy v5.1.1 -> v6.0.3
* fix: restore i18n, UI, and docs work undone by 0e9218bd68
Commit 0e9218bd68 silently reverted the following work alongside its stated
sync/client-id changes. Files where later master commits (fec7b25f23, etc.)
already re-applied the reverted work are intentionally left untouched.
Restored:
- #7232 docs/long-term-plans/location-based-reminders.md (513 lines)
- #7199 Romanian i18n phase 3 (ro.json + ro-md.json, ~1168 lines)
- #7049 Polish translation improvements
- #7143 planner component styling (4 scss files)
- #7211 add-task-bar preserve time estimate when typing title
- #7208 task.reducer roll-up estimates for added subtasks
- #6767 focus-mode pomodoro reset button
- #7205 plugin-dev github-issue-provider TOKEN description
- #7231 mobile-bottom-nav FAB fix
- a4fe03272 iOS keyboard accessory bar (global-theme + dialog-fullscreen-markdown)
- 309670db3 ShortSyntaxEffects undefined guard
- 667a7986f Dropbox PKCE auth comment/behavior
* fix(electron): restore electron + e2e work undone by 0e9218bd68
Commit 0e9218bd68 silently reverted the following electron/e2e work.
Files already re-fixed by later master commits are left as-is:
- e2e/tests/sync/supersync-archive-conflict.spec.ts (de33234976 + 191d129ff3)
Restored:
- e8a3e156eb fix(electron): Linux autostart IDB backing-store recovery
(re-adds electron/clear-stale-idb-locks.ts + start-app.ts wiring)
- 5ce78a5b63 fix(electron): macOS Cmd+Q / Dock > Quit hang
(setIsQuiting + before-quit delegate to close-handler)
- 46e0fa2d01 fix(sync): FILE_SYNC_LIST_FILES IPC contract
(electronAPI.d.ts + local-file-sync + preload + ipc-events)
- ea1ef16307 fix(android): session-only SAF permissions on OEM devices
- 8865dc0a50 test(e2e): supersync parallel-worker stampede guard
(SUPERSYNC_SERVER_HEALTHY env-var fallback + goto retry loop)
- af7c7687e2 test(e2e): block WS-triggered downloads in non-WS specs
- 265b44db5d test(e2e): premature waitForURL on daily-summary
(this is literally the fix the bad commit's message claimed to add)
Conflict resolutions:
- e2e/utils/supersync-helpers.ts: kept the refined getDoneTaskElement
checks from d64014d086 (later than c558bcab5e) while restoring the
goto retry loop from 8865dc0a50.
- electron/start-app.ts: unioned imports (setIsQuiting +
clearStaleLevelDbLocks from theirs, fs from ours).
* fix(sync): restore sync-core work undone by 0e9218bd68
Commit 0e9218bd68 silently reverted parts of several sync fixes. Most
sync-core reverts have already been re-addressed differently on master
by later commits (1f5184f6e7, 05cd875dd6, 09f5ced2c9, 7df43358ab,
d9158d6adb, 32dbc95ed9, 8c3b08e016, f89fe1ebc3) — those files are
intentionally left untouched to avoid reverting master's newer work.
This PR restores only the pieces that are genuinely still missing:
- e8a3e156eb fix(electron): IDB backing-store autoreload (in-app piece)
operation-log-hydrator.service.ts + .spec.ts (the electron/clear-
stale-idb-locks.ts piece was restored in the prior commit)
Plus three low-risk documentation/cleanup restorations:
- operation-sync.util.ts — add "Nextcloud" to isFileBasedProvider JSDoc
- dropbox.ts — restore improved _getRedirectUri JSDoc (667a7986fb)
- dialog-get-and-enter-auth-code.component.ts — restore isNativePlatform
comment explaining why manual code entry flow is used (667a7986fb)
- file-adapter.interface.ts — remove stray "// NEW" comment (46e0fa2d01)
Intentionally NOT restored (master's newer work covers or supersedes):
- sync-trigger.service.ts / sync.effects.ts (05cd875dd6)
- sync-wrapper.service.ts (1f5184f6e7)
- sync-errors.ts (1f5184f6e7 re-added LegacySyncFormatDetectedError)
- file-based-sync-adapter.service.ts + spec (1f5184f6e7 + d9158d6adb)
- file-based-sync.types.ts (1f5184f6e7)
- operation-log.const.ts (32dbc95ed9 bumped IDB_OPEN_RETRIES to 5)
- dialog-sync-initial-cfg.component.ts (f89fe1ebc3)
* docs(research): add Snap + Wayland GPU init failure research
Synthesizes root cause (Mesa ABI drift in gnome-42-2204), scope,
peer-app consensus, and ranked options. Recommends widening the
existing Snap-gated --ozone-platform=x11 guard in start-app.ts to
cover Snap + Wayland sessions, with core24 + gpu-2404 migration
as the long-term fix.
https://claude.ai/code/session_01Hu6EP7Xux9JRvGGGKpBfwm
* docs(research): correct electron timing and soften unverified peer claims
* fix(electron): force X11 on Snap Wayland to avoid Mesa ABI drift
The existing Snap-only X11 fallback triggered only when $SNAP/gnome-platform
was empty. In practice gnome-platform is populated but its Mesa can drift
out of ABI sync with Electron's bundled libgbm, producing
"Failed to get system egl display" and a GPU process respawn loop on
Wayland sessions. Widen the guard to also fire when the session is
Wayland (XDG_SESSION_TYPE=wayland or WAYLAND_DISPLAY set). X11/GLX
avoids the failing Wayland EGL init path entirely while preserving
hardware acceleration.
Refs: electron-builder#9452, super-productivity#5672,
forum.snapcraft.io #40975, #49173https://claude.ai/code/session_01JNxazJmDhpMp9UYV6SqnBG
* refactor(electron): address review feedback on Snap X11 force
- Handle space-separated `--ozone-platform wayland` override in addition
to the `=` form; tighten the argv scan to reject accidental substring
matches (e.g., arg values that happen to contain `--ozone-platform=`).
- Enrich the log line with raw XDG_SESSION_TYPE and WAYLAND_DISPLAY
values to make user-reported triage faster.
- Tighten comment: correct "X11/GLX" to "X11 ozone backend" (modern
Chromium uses EGL on X11), clarify the root cause as Mesa version
drift rather than bundled-library ABI drift, broaden the loss list
beyond fractional scaling, and move issue references to the commit
body per CLAUDE.md guidance.
Follow-ups flagged but not addressed here: possible interaction with
`allowNativeWayland: true` in electron-builder.yaml, GPU cache keyed
by effective ozone platform, and the longer-term core24 migration
already in docs/long-term-plans/electron-upgrade-to-v40.md.
Refs: electron-builder#9452, super-productivity#5672,
forum.snapcraft.io #40975, #49173https://claude.ai/code/session_01JNxazJmDhpMp9UYV6SqnBG
* docs(research): verify peer-app claims, correct SP electron timing
---------
Co-authored-by: Claude <noreply@anthropic.com>
* chore: update electron to v41.1.1
* feat(start-app): clear GPU cache on Electron version change for Linux
* fix(window-decorators): disable custom window title bar for GNOME
This fixes some issues when moving and resizing the window
* refactor: Migrate url.format() (DEP0116) to file:// path approach
* refactor(start-app): remove deprecated protocol.registerFileProtocol in start-app.ts
* feat(electron-builder): update gnome content snap to gnome-42-2204 for improved compatibility and update flatpak permissions
* fix(window-decorators): improve handling of custom window title bar for GNOME
* feat(start-app): implement fallback to X11 in Snap if gnome-42-2204 runtime is unavailable
* chore: update electron to v41.1.1
* chore(package-lock): remove unused dependencies from package-lock.json
* fix(electron): move snap ozone-platform switch before app ready event
app.commandLine.appendSwitch() must be called before Chromium
initializes — after the ready event fires the GPU backend is already
running and the switch is a no-op. Move the Snap X11 fallback (defense-
in-depth for missing gnome-42-2204 runtime) to run synchronously at
startup, alongside the existing gtk-version and speech-dispatcher
switches.
* fix(electron): align IS_GNOME_DESKTOP detection with preload.ts logic
The original implementation only matched Ubuntu's GNOME session
(XDG_CURRENT_DESKTOP contains both 'gnome' AND 'ubuntu'), missing plain
GNOME on Fedora, Arch, etc. The preload.ts isGnomeDesktop() already
used the correct approach: check four environment variables with OR
logic. Align common.const.ts to match, preventing a split-brain where
the main process and renderer disagree on whether the user is on GNOME
— which would produce no title bar at all on non-Ubuntu GNOME desktops.
* fix(electron): restore v41 bump lost in merge and gate title-bar toggle
- Re-apply electron 41.2.0 + minimatch 10.2.5 override (master's 0e9218bd
reverted the dependabot bump back to 37.10.3 while this branch's
merge-base still contained 41.2.0, so the pre-merge diff was empty).
- Regenerate root package-lock.json accordingly.
- Drop unrelated esbuild additions from plugin-dev sub-lockfiles.
- misc-settings-form: gate isUseCustomWindowTitleBar on IS_ELECTRON &&
!IS_GNOME_DESKTOP so the toggle does not appear in the web/PWA build.
---------
Co-authored-by: Johannes Millan <johannes.millan@gmail.com>
- Separate JsonParseError and SyncDataCorruptedError handlers in sync wrapper
- Handle corrupted remote data gracefully instead of throwing
- Add getOrGenerateClientId() as unified entry point, eliminating dual injection
of ClientIdService + CLIENT_ID_PROVIDER in snapshot-upload, file-based-encryption,
and sync-hydration services
- Use crypto.getRandomValues() instead of Math.random() for client ID generation
- Warn user when stored client ID is invalid and must be regenerated
- Fix flaky e2e supersync tests (premature waitForURL match on daily-summary URL)
Three-layer defense for issue #7191 (Flatpak/Snap/autostart startup race):
1. Increase IDB_OPEN_RETRIES 3→4 (total retry window 3.5s→7.5s) to give
OS/Flatpak storage more time to initialize before giving up.
2. Silent auto-reload on first backing-store error: instead of a blocking
alert that requires user interaction before the reload (bad for autostart
where nobody is watching), reload immediately and only show the manual
recovery dialog if the reload does not fix the problem.
3. Clear stale LevelDB LOCK files in the Electron main process before the
renderer opens IndexedDB. Orphaned locks from unclean session shutdowns
block the backing store open; safe to remove because requestSingleInstanceLock()
already confirmed we are the only running instance.
Fixes https://github.com/super-productivity/super-productivity/issues/7191
On macOS, native quit fires before-quit BEFORE window close events. The
before-quit handler was calling ipcMain.removeAllListeners(), which destroyed
the BEFORE_CLOSE_DONE IPC listener. The close handler then sent NOTIFY_ON_CLOSE
to the renderer (for sync/finish-day callbacks), the renderer responded with
BEFORE_CLOSE_DONE, but the listener was gone — causing a permanent hang.
Fix: intercept before-quit when isQuiting=false and delegate to win.close(),
which re-enters the existing close handler and runs before-close callbacks
normally. Move ipcMain.removeAllListeners() to will-quit, which fires after
all windows are closed and all IPC flows are guaranteed complete.
Also improves close event log to include pending before-close handler IDs.
When a full-state op (SYNC_IMPORT/BACKUP_IMPORT) arrived from another
client, mergeRemoteOpClocks reset the local clock to a "minimal" form
but only preserved the current client's counter from the incoming op's
clock. If the current client had issued ops (e.g. GLOBAL_CONFIG) not
reflected in the incoming full-state op's clock, its counter was dropped,
causing subsequent ops to reuse the same counter value. Downstream clients
then saw these ops as EQUAL (duplicate) and skipped them silently.
Fix: take max(mergedClock[clientId], currentClock[clientId]) when
rebuilding the clock after a full-state op reset.
Also add __SP_E2E_BLOCK_WS_DOWNLOAD flag to WsTriggeredDownloadService
to allow E2E tests to block automatic WS-triggered downloads during
concurrent conflict scenarios.
Fix archive conflict test by blocking WS downloads on Client A during
the concurrent edit phase so it doesn't auto-receive B's rename via
WebSocket before archiving (restoring the intended conflict scenario).
Fix LWW singleton test to assert convergence rather than specific winner.
Fix renameTask helper to avoid Playwright/Angular re-render races.
Fix shepherd.js import paths that broke the Angular dev server build.
electron-dl v4 is a pure ESM package. Since the Electron main process
is compiled with TypeScript "module": "commonjs", static import() gets
transformed to require() which cannot load ESM. Use Function() to
preserve a real dynamic import() at runtime.
Rename the "overlay indicator" feature to "task widget" across the
entire codebase for clearer naming. Includes file renames, config key
migration, IPC event rename, and translation key updates.
Migration handles old persisted data from both the `overlayIndicator`
config key and the deprecated `misc.isOverlayIndicatorEnabled` field.
The `taskWidget` type is made optional in GlobalConfigState to prevent
Typia auto-fix from overwriting migrated values during upgrade.
Enable webSecurity (same-origin policy) since CORS is already handled at
the session level via onBeforeSendHeaders and onHeadersReceived. Add
setPermissionRequestHandler to deny unnecessary permissions (webcam,
microphone, geolocation). Improve certificate-error logging with warn
level and URL context.
Add OAuth support for plugins with PKCE, token persistence in local-only
IndexedDB, and Electron/web redirect handling. Extend two-way sync with
dueDay, dueWithTime, and timeEstimate field mappings including mutually
exclusive field clearing. Support remote issue deletion on task delete
and remote deletion detection during polling.
Add Google Calendar plugin (disabled from bundled builds pending legal
review) as a development reference for the plugin OAuth and two-way
sync APIs.
Additional fixes:
- Restore accidentally deleted focus-mode translation keys
- Remove full Task[] from deleteTasks action to prevent op-log bloat
- Add dueDay/deadlineDay format validation guards (#6908)
- Fix Android reminder alarm cancel intent matching
- Validate dueDay format to prevent false overdue from corrupted data
- Fix PKCE test expectation after utility consolidation
Suppress noisy "EXITING due to failed single instance lock" message when
a second instance is launched via xdg-open for protocol URL handling.
Register superproductivity:// as a MIME type handler in the .desktop file
so Linux users no longer need manual xdg-settings configuration.
Remove duplicate requestSingleInstanceLock() call in start-app.ts.
Closes#173
- Gate blur() to Windows only (not supported on Wayland, limited on macOS)
- Add setTimeout delay in ready-to-show to match showOrFocus pattern
- Add webContents.focus() in ready-to-show for renderer-level focus
- Add isDestroyed() guards in setTimeout callbacks
- Only restore focus on resume/unlock if window was visible before
suspend/lock to avoid surfacing hidden/minimized windows
On Windows, BrowserWindow.show() can silently fail to acquire keyboard
focus after a system reboot (electron#20464). This leaves the window in
a "phantom focus" state where clicks work but typing does not.
- Add blur()+focus() cycle after show() in ready-to-show handler
- Add webContents.focus() in showOrFocus() delayed callback
- Add showOrFocus() call in resume/unlock-screen power monitor handlers
Fixes#6663
requestSingleInstanceLock() creates a Unix domain socket blocked by the
macOS App Store sandbox, causing the app to silently fail to start (#6275).
macOS natively prevents multiple app launches, so the lock is unnecessary.
Remove the code that forced X11 mode on Linux via --ozone-platform=x11.
This was causing the snap package to crash on Wayland systems due to
GPU driver mismatches between the bundled Mesa drivers and host system.
The IdleTimeHandler already has proper fallbacks for Wayland environments
(gnomeDBus, xprintidle, loginctl), so forcing X11 is not necessary.
Closes#5663
Upgrade Electron from 37.7.0 to 39.2.5. Since Electron 38+ defaults to
Wayland via --ozone-platform=auto, force X11 on Linux to ensure reliable
idle detection (#1443) and global shortcuts. Users can opt-in to Wayland
with --ozone-platform=wayland or --force-wayland flags.
Add comprehensive logging to help diagnose why idle detection is not
triggering properly when the system suspends/resumes, particularly on
Wayland systems where powerMonitor events may not fire reliably.
The logging covers:
- Power monitor suspend/resume/lock/unlock events
- Idle time calculations on system wake
- Whether idle messages are sent to frontend based on thresholds
This will help identify if:
1. PowerMonitor events are firing on the user's KDE Wayland system
2. Idle time is being calculated correctly during sleep periods
3. Messages are being sent but not processed by the frontend
Related to issue #4734 where time tracking continues during sleep mode.
