The Library of Congress Just Made SQLite an Official Preservation Format

Imagine your database file getting the same museum-grade treatment as a Gutenberg Bible. That’s roughly what happened last week: the U.S. Library of Congress quietly added SQLite to its Recommended Storage Formats, slotting a database engine right next to CSV, JSON, and PDF/A. As far as anyone can tell, it’s the first time a structured-data container has earned that designation.

What “Recommended Storage Format” actually means

The Library of Congress ingests tens of petabytes of digital material every year. The uncomfortable question hanging over all of it: will anyone be able to open these files in 2126? The Recommended Format Statement is their answer.

The bar is high. A format must have a publicly documented spec, no vendor lock-in, broad adoption, and ideally be self-documenting. That’s why PDF/A, CSV, FLAC, and TIFF made the cut — and why Microsoft’s .doc never did.

SQLite joining that list is a bigger deal than it looks. Until now, the recommended formats were essentially documents. SQLite is the first time a container for structured data has been formally blessed for long-term preservation.

Why SQLite, of all things

SQLite started in 1989 — older than most of the web. One file, no external dependencies, source code in the public domain. There’s literally no license anyone can leverage against you in 2080.

It’s also already everywhere. Your iPhone’s contacts, Android’s SMS database, Chrome’s bookmarks, even flight data recorders all run on it. Estimates put the number of active SQLite databases worldwide at over one trillion. If the criterion is “broad adoption,” nothing else in the data-format world comes close.

The clincher: the SQLite team has publicly committed to maintaining backward file-format compatibility through 2050. A .sqlite file you write today is guaranteed to open three decades from now. For an institution that thinks in centuries, that kind of commitment is rare currency.

The bigger shift: a database is now a document

Here’s the part worth sitting with. For decades, the reflex of digital preservation has been “dump the data to CSV and call it a day.” Databases were treated as volatile systems — useful for queries, not for archives.

SQLite breaks that frame. No server, no config, no daemon — just a file. The Library of Congress is effectively saying: if your data has relationships and schema, preserve the relationships and schema, not just the rows.

Think of it like Excel. Export to CSV and your formulas vanish. Dump a relational database to CSV and you lose foreign keys, indexes, constraints, the whole web of meaning. A SQLite file keeps all of it intact.

How the developer world is reading it

Hacker News framed it as the payoff of D. Richard Hipp’s 30-year bet. But the more interesting signal is institutional: when a body as conservative as the Library of Congress moves, it usually means the underlying trend — government agencies, scientific consortia, and research labs already using SQLite as an interchange format — has crossed a tipping point.

There’s a fun footnote, too. SQLite was originally built for guided-missile destroyers in the U.S. Navy. A database engineered to survive combat conditions is now being deployed to safeguard humanity’s digital memory. Different kind of mission, same design philosophy.

The practical takeaway

If your company holds data that someone will need to open in 2046, this decision is more than archival trivia. Locking records inside a niche vendor’s proprietary format, or trusting a SaaS export menu that may not exist in five years, just got harder to justify. A plain SQLite file is now, quite literally, library-grade.

A database has been promoted from volatile infrastructure to permanent record. So where is your important data sitting tonight, and in what shape? Worth a look before the next migration forces the question for you.