Work with timestamps, IDs and hashes without false confidence

Confirm the timestamp unit and timezone

Unix timestamps are usually seconds since 1970-01-01 UTC, but many APIs send milliseconds. A value that turns into a date thousands of years away is often a seconds-versus-milliseconds mistake. Check the producing system’s documentation before changing the number.

Convert in both directions when debugging. Keep timezone separate from the timestamp itself: the instant is UTC-based, while this tool shows both UTC and your device’s local timezone.

Do not confuse unique IDs with secrets

A UUID is useful as a locally generated identifier for records, filenames or test data. It is not a password and should not be used as an access token unless a system specifically designs for that use.

For a password, generate a long random value and store it in a password manager. Randomness matters more than a clever pattern, and a password generator cannot make reuse across sites safe.

Use hashes to compare bytes, not to hide data

A hash gives a fixed fingerprint of exact input. Compare the same algorithm on the same file or text to check whether they match; a changed byte is expected to produce a different digest. It does not tell you where a change happened or prove who created a file.

Hashes are not encryption. Do not hash a password yourself as a storage scheme, and do not assume a short or predictable value is protected merely because you have its hash.

Reproduce a log-record check

If a test record contains created_at: 1704067200, select Seconds in Timestamp Converter: the UTC result is 2024-01-01T00:00:00.000Z. The equivalent millisecond value is 1704067200000. Check the UTC result rather than comparing a local clock display from a colleague in another timezone.

For a record ID, v4 supplies random identifiers; v7 embeds a millisecond timestamp. This generator fills the remaining v7 bits randomly, so IDs made within the same millisecond are not guaranteed to sort in creation order. An ID’s timestamp is not proof that the event happened then.

To compare a payload fingerprint, hash the text abc with SHA-256 and compare it with the same text followed by a newline: the digests differ. In file mode the full contents are read into memory, with a 256 MiB limit. Obtain an expected digest from a trusted source; replacing both a file and its checksum would defeat this comparison.