The 2038 Problem

On 19 January 2038, at 03:14:07 UTC, time will overflow.

Not time itself — but the timestamp format used by billions of devices, systems, and protocols around the world.

This is not science fiction. It’s not hypothetical. It’s not even “in the future.”
It is already embedded in critical infrastructure — shipping now, running now, failing silently in test environments now.

The year 2038 marks a boundary not just in code, but in trust.
— Threshold Continuity Alliance


What’s at Risk?

The Unix Epoch — the format many computers use to measure time — was born in 1970. It assumes that time is stored as a 32-bit signed integer. That format overflows in early 2038, like an odometer rolling past its maximum value.

When that happens:

Worst of all?
Many of these failures may not be visible until it’s too late.


Why It’s Worse Than Y2K

Y2K was about formatting dates. The 2038 problem is about how systems fundamentally understand time.

It’s not just a software issue. It’s a supply chain, governance, and epistemic issue.
Who will trust a system whose timestamps can’t be verified?


What Can Be Done?

There is still time — but not much.

🧪 Technically:

Fixing time is fixing trust.

📜 Policy & Procurement:


What We’re Building

Long Now EU exists to carry forward the tools, rituals, and engineering needed to preserve coherence across this boundary.

We are collaborating with:

A public GitHub toolkit will include:

Join us. Build what the future needs.


Less Than 50 Quarters Remain.

Every unpatched bug is a future failure.
Every timestamp overflow is a breach of trust.
Every resilient system begins with a human who refuses to stay silent.

The clock is ticking. But time is still ours to shape.
We are not just counting down — we are building forward.

📬 Contact us if you want to help.