Self hosted unrot process

This server is based on the Netlify Edge functions. So if you're also hosting on Netlify, you should be able to use this source code or a variation of it.

The full source code is available on github: remy/unrot.link.

The redirect process can be seen in the interactive "how it works" page.

A simplified implementation of the backend has been documented over on Remy Sharp's blog post.

Jeremy Keith has written a version that runs using PHP (seen under the "Link rot" heading) (original blog post).

If you plan to design your own version, please consider posting a pull request to the repo so that others can benefit from your work.

API

This domain, https://unrot.link is the API endpoint for the service. It accepts a single GET request with the following parameters:

• url (required) - the URL to check for link rot
• date - the timestamp to use when checking for a web.archive.org copy of the URL. This should be in the format YYYY-MM-DD HH:MM:SS (e.g. 2019-01-01 00:00:00 for 1st January 2019 at midnight). If not provided, the latest available copy will be used. Note that the time is not required.
• origin-match - a boolean to return a web.archive.org link if the URL tries to redirect to a different origin. This is useful if you want to ensure that the link is still pointing to the same domain, but potentially problematic if a domain redirects to an intermediary (as for example yahoo does). Defaults to false.
• wayback - a boolean to force unrot.link to always return the web.archive.org copy of the URL, even if it's not a 404. Defaults to false.
• timeout - the number of milliseconds to wait for a response from the given url before assuming a failure and moving onto requesting the page from the web.archive.org API. Defaults to 2000 (2 seconds).