The Memento Protocol (RFC 7089)
Time travel for the web: Understanding the HTTP framework for time-based access to resource states.
What is Memento?
The web was built for the present. HTTP standardizes how to get a resource right now. The Memento protocol (standardized as RFC 7089) extends HTTP to answer the question: "What did this resource look like on a specific date?"
How It Works
Memento introduces new HTTP headers. When a client (like a browser extension) wants to view an old version of a page, it sends a standard GET request to the original URL, but adds an Accept-Datetime header specifying the desired timestamp.
If the server supports Memento (or if the request is routed through an archive aggregator), it replies with a 302 Redirect to a WARC record located in an archive like the Wayback Machine or archive.today, representing the closest available capture to that date.
Combating Link Rot
Memento is the ultimate technical solution to Link Rot. If implemented at the browser level, a user clicking a dead link would automatically be routed to the closest archived version without manual intervention.
Memento Aggregation (Time Travel)
Because Memento is a standard protocol, it allows for aggregation. Services like the Los Alamos National Laboratory's Time Travel service query multiple archives simultaneously using Memento, similar to our Multi-Archive Tool, and return the best result.
Memento FAQ
Do regular web servers support Memento?
Very rarely. It requires the server to maintain historical versions of all resources, which is computationally expensive. It is primarily implemented by web archives and a few wikis.