Raw capture — raw/lindy-effect-in-software.md

url: https://www.clemsau.com/posts/the-lindy-effect-in-software/ title: The Lindy Effect in Software author: Clément Sauvage date_fetched: 2026-07-08 date_published: 2023-06-17 site: clemsau.com

The Lindy Effect in Software

Author: Clément Sauvage Published: 17 June 2023 Reading time: 4 minutes Tags: Software, Engineering

Overview

The article introduces the Lindy effect — a concept suggesting that the longer something has survived, the longer it can be expected to continue surviving. Sauvage applies this principle, originally observed in fashion and books, to the world of software technology.

What Does the Lindy Effect in Software Mean?

The author proposes that a technology's longevity may signal its robustness, making it a safer bet than newer, trendier alternatives. He acknowledges that software is young relative to other domains and that evolution (e.g., Apple Silicon) is welcome — but that "groundbreaking benchmark" results alone shouldn't trigger a full migration.

Sauvage references the "choose boring tech" philosophy and lists five benefits of established technologies:

  1. Stability and Reliability — Older systems have had bugs and vulnerabilities addressed over time.
  2. Mature Ecosystem — Extensive documentation, libraries, and community support exist.
  3. Predictable Performance — These technologies are "battle-tested in numerous real-world scenarios."
  4. Industry Acceptance — Skilled professionals are easier to find.
  5. Reduced Risk — Novelty introduces unknowns like undiscovered bugs or uncertain long-term support.

He contrasts Go versus Java for enterprise backends, noting he personally favors Go but understands why large companies "play the safest bet" with Java given decades of institutional experience. SQL (since 1989) and C (since 1972) are cited as examples of technologies whose long track records suggest continued relevance.

Applying the Lindy Effect to Software Engineering

Sauvage proposes four practical steps for engineers seeking balance between new and time-tested tools:

  1. Prudent Adoption — Carefully evaluate new tech before committing.
  2. Stick to Proven Foundations — Use established tech for core components.
  3. Plan for Longevity — Design architectures with maintainability and sustainability in mind.
  4. Embrace Evolution, Not Revolution — Favor incremental improvement over constant rewrites.

Link to original: The Lindy effect in software