The boring-technology argument for long-lived systems

Every new framework promises a breakthrough. Some deliver. But for a system you will own for years, the question is not “what’s the newest” — it’s “what will still be supported, documented, and understood by the next engineer in five years?”

We call it the boring-technology principle: for the load-bearing parts, choose the option with the deepest ecosystem, the best documentation, and the least surprise. Postgres over the flashiest database. Boredom over novelty. The goal is to minimize the cost of the next hire understanding the system.

Where we do take risks

We’re not purists. For edge features — an AI integration, a prototype, a time-boxed experiment — new tools are exactly right. The principle is about the spine of the system, not the leaves.

If your team keeps rewriting the same core three times a year, that’s a technology-choice problem, not a discipline problem.