The Record That Outlives Us

The Problem

A record maintained by a company is only as durable as the company, and companies are not durable.

This is uncomfortable to design around because it requires assuming your own absence at the point where you are most invested. It is also the only assumption that makes the record worth anything, since the whole promise of an archive is that it holds after the moment it was made in.

The Idea

Anything that requires us to be reachable was disqualified.

That single filter removed most of what had been planned. A verification service that phones home. A dynamic certificate. An account that has to be logged into. Each of them works beautifully while somebody is paying the hosting bill and becomes a dead end at the exact moment the record is most needed.

Building The Archive

What survived the filter is small: a physical mark on the object, a physical document issued with it, and a digital record that mirrors both without being required by either.

The mirror is the part most people assume is the archive. It is not. It is the convenient copy. If it disappeared tomorrow, the object would still carry its number and the document would still carry its history, and the record would be inconvenient rather than lost.

Building it in that order — object, document, then database — was a constant fight against convenience, because the database is the only one of the three that can be changed in an afternoon.

Challenges

The main challenge is that this design has no visible payoff during the period when you are building it.

Everything it protects against happens later, to someone else, and there is no feedback loop telling you it was worth doing. The alternative design looks better in every demonstration, right up until the point of failure that nobody demonstrates.

What Changed

Once the standard was set, it began deciding things far outside its original scope.

Whether a field belongs on the certificate. Whether a repair gets recorded on the object or only in the file. Whether a piece of information is worth capturing at all, given that capturing it creates an obligation to keep it readable for decades.

Most proposals fail that last test, and the archive is smaller and more durable for it.