unrot•link

This interactive document explains the steps that the redirection server takes to determine the best URL to use.

Request options

URL to check: wayback? skip 404 check and always use the Internet Archive origin-match? if the target url responds with a different origin, use Internet Archive date? timeout (ms)?

After being submitted, the parameters go through validation before the redirect logic starts.

Attempt to connect to the origin with a timeout of 200ms.

If the timeout triggers, force the use of the Internet Archive and don't try to check if the url is responding.

Decide what would happen next in the logic.

Do you want the origin connection succeed? yes no

Source

Fetch a HEAD request for the url
If the HEAD returns a 405, request again using GET
Allow for timeout milliseconds for a response

See how the script responds to different fetches

Source

If match-origin is set, check the origin of the redirect (from the target url). If it doesn't match, then use the Internet Archive

If the url was redirected to the root of the server (and the url wasn't the root) then treat as a 404 and use the Internet Archive

Otherwise, treat as legitimate and do not use Internet Archive

What kind of redirect was it?

Origin change Path change to root Legitimate

Source

Redirect the user to the original url with a 301 "Moved permanently" status code.

End of logic

Source

Use the wayback CDX API to query for status code 200 and attempt to collect the latest entry if no date given, and first entry if a date is given.

Source

If the Internet Archive request is successful, redirect the user to the stored URL with 301 "Moved permanently" status.

If it fails, or there's no data for the given url, then the user is redirected to the original url with 302 "Found" status.

Given your parameters, the result would redirect to:

checking…

End of logic

Source

How unrot∙link works