Skip to main content

Practical guidance and evidence

Accuracy & Reliability

Find practical answers about clock sources, sound, browser lifecycle, and local data before relying on a tool. The technical registry below maps public claims to checked-in source, repeatable tests, and explicit limits.

Practical answers

Which clock and time-zone source does Best Online Clock use?

The main clock reads the current time exposed by your browser and device system clock, then refreshes on an aligned one-second ticker. City pages use checked-in IANA time-zone identifiers with the browser's Intl date/time formatter.

Enable automatic date, time, and time-zone settings on the device when available. This is not atomic-clock synchronization or a calibration certificate. A wrong or stale device clock, browser time-zone data, or changed government time-zone rule can affect the displayed result.

How should I test sound, volume, and mute settings?

Test playback with a direct user action before relying on an alert. For a timer, open Timer Settings > Sound, choose a completion sound, and press its play preview control. For an alarm, open Alarm Settings, choose a sound, and use the play alarm sound preview; the existing Test Alarm control can test the alarm flow as well.

Check the device and operating-system volume, the selected output, and that the browser tab or site is not muted. If a timer completion reports a blocked or failed sound, select Retry timer sound, then check the connection and volume or choose another sound. For an alarm preview, press its play control again after correcting the issue. Browser autoplay rules, permission state, and device policies can still block or change playback.

Will a timer or alarm wake me in the background or on the lock screen?

A running countdown keeps an absolute deadline and recalculates its remaining time when the page resumes, but the tab still needs to remain open. Background-tab throttling, suspended audio, battery-saving behavior, and lock-screen policies can affect what the browser is able to do.

Daily alarms are browser-page features, not an operating-system or account service. Fullscreen can request the Screen Wake Lock API when the browser and device support it, but the request may be denied or released. There is no guaranteed background or lock-screen wake-up, so test the behavior on the device and browser you plan to use. For an important reminder, set a separate device alarm as a backup.

What is saved in my browser, and does it sync to another device?

Tool settings, alarms, selected cities, and uploaded background data are stored in the current browser using localStorage and IndexedDB. There is no account sync; clearing browser storage or changing browsers or devices can remove or separate that local data.

This describes local tool state only. The Privacy Policy separately explains contact submissions, analytics, advertising, Clarity, and other network services. Local tool storage does not mean every network request stays local.

Technical evidence, review method, and limits (g2_facts_v1)

Fact schema g2_facts_v1 · reviewed

How claims are reviewed

  1. 1. Trace the behavior. Identify the state transition, storage boundary, browser API, and user-visible result in source.
  2. 2. Fix the clock. Unit tests use fixed instants for countdown, alarm occurrence, DST, date rollover, and time-zone calculations.
  3. 3. Exercise the page. Browser tests cover task completion, keyboard access, responsive layouts, reload/resume behavior, and error states.
  4. 4. State the boundary. Claims are rejected or narrowed when audio, notifications, background execution, storage, or device support cannot be guaranteed.

This lower-level registry keeps public statements tied to implementation and automated checks. Evidence paths are repository paths, and each fact retains its limitation rather than implying a guarantee beyond the tested behavior.

Timer deadlines

Running countdowns use an absolute saved deadline and recalculate the remaining time when the page resumes.

Checked-in evidence

  • src/features/timer/core/countdown.ts
  • src/features/timer/core/countdown.test.ts
  • e2e/timer-accuracy.spec.ts

Limit

The tab must remain open. Browser throttling, audio, notifications, and Wake Lock behavior still depend on the browser and device.

Daily browser alarms

Permanent alarms repeat daily at the selected device-local time; snooze choices are 1, 5, or 10 minutes.

Checked-in evidence

  • src/features/alarm/core/alarm.ts
  • src/features/alarm/core/alarm-reliability.ts
  • e2e/g2-alarm-presets.spec.ts

Limit

An alarm is not a background account service. Keep the page open and test sound before relying on it.

Stopwatch timing

The stopwatch uses the browser monotonic performance clock, records laps and splits, and exports the current session as CSV.

Checked-in evidence

  • src/features/stopwatch/core/stopwatch.ts
  • src/features/stopwatch/core/stopwatch.test.ts
  • e2e/stopwatch-accuracy.spec.ts

Limit

Displayed precision is not a calibration certificate and the tool is not intended for safety-critical or regulated timing.

World-time calculation

City clocks use checked-in IANA time-zone identifiers with the browser Intl API; Manila is Asia/Manila, UTC+8, with no current DST rule.

Checked-in evidence

  • src/data/cities.json
  • src/features/world-time/core/timezone.ts
  • src/features/world-time/core/timezone.test.ts

Limit

Time-zone rules come from the browser runtime and can change when governments or runtime data change.

Local tool data

Tool settings, alarms, selected cities, and uploaded background references are stored in the current browser without requiring an account.

Checked-in evidence

  • src/lib/settings-storage.ts
  • src/lib/background-image-storage.ts
  • src/features/alarm/alarm-client.tsx
  • src/features/world-time/components/multi-city-clock.tsx

Limit

Clearing browser storage or changing devices can remove or separate that local data.

What this page does not claim

Automated Chromium, Firefox, and WebKit checks are useful regression evidence, but they do not replace every physical device, operating-system policy, real Safari build, or future browser version. Real-user performance samples are reported only after their stated observation window and sample gate.