Stacey on Software

A 45-year-old computer learns to invent computer names

My Tandy Color Computer from 1981 starts from random numbers and, a minute later, invents computer names that never existed. Post one of five on how it does that.

What Sixteen Models Did With the Same Eight Principles

313 trials across sixteen models, one Rust scenario, the same eight engineering intents. Lift ranged from 0.00 to 0.89 depending only on which model read them — and the strongest model in the set was nowhere near the top.

The Spec for the Work Isn't the Spec for the Thing

I looked at GitHub's spec-kit when it launched a year ago, found it incomplete, and carried on with my own line of work. Coming back to it, I can finally name what was missing: it builds a project spec, and I keep wanting a product spec.

Modelling Engineering Intent Made My Guidance Measurable

I'd been trying to test my agent guidance for a couple of years, and every attempt died in the same place. What changed wasn't the harness — it was modelling engineering intent as structured records, which made the guidance specific enough to measure at all.

A Leaf on the Wind, at Machine Speed

Agents have changed the economics of producing and repairing software. They haven't changed the physics of entropy.

When Intent Moves Faster Than Code

Agents build code so fast that intent has become the bottleneck. And in a large codebase, the friction between fluid intent and rigid structure becomes the thing that actually limits how fast you can move.

Seven Coding Agents, One Space Invaders Clone

Anthropic is splitting their subscription on June 15, and unattended Claude Code is off my menu. I ran the same Bevy build through seven coding-agent profiles to see what could replace it. Three of them produced playable games. Four shipped binaries that crash on launch. The cost spread between the working ones is six-to-one, and the gameplay is indistinguishable.

Every Codebase Is an Uncompiled Knowledge Base

Karpathy's LLM Knowledge Base architecture is a powerful contribution, but it draws a false boundary between code and knowledge. Your codebase is the largest uncompiled knowledge base in your organization — and agents can now extract what's inside it.

Code Is Knowledge

Code isn't just instructions for machines. It's a record of how an organization thinks — about customers, domains, risks, and trade-offs. We've always known this. Now we have tools that can make it explicit.

Intent in Motion

A product's source code is a point-in-time snapshot of intent in motion. When an organization protects the artifact more than the community that gave it meaning, you learn what they actually believe.

GenAI Didn't Kill Rigor — It Moved It

The old bottleneck was implementation. The new bottleneck is ambiguity. And the rigor that used to live in coordination now needs to live in specification, context, and evaluation.

Know Your Principles

Comprehension debt isn't just about AI-generated code you don't understand. It's about not being able to articulate the principles that make your codebase navigable in the first place.

Finding Your True Worth

Your value isn't in typing code—it's in what you know. In a GenAI world, that tacit knowledge is more precious than ever.

The Agent That Won't Defect

Why agentic coding might be our best defense against the social dynamics that create tech debt

The Hacker Manifesto at Forty

Forty years ago, a frustrated teenager wrote a manifesto from jail. It still resonates — and still defines what it means to be a hacker.

Building Consistent AI Imagery with Structured Prompts

How I built a reusable system for generating consistent blog imagery using JSON schemas, the OpenAI image API, and Claude Code skills.

We Fixed the Glitch

Part one of four on Russell Ackoff's problem treatments. Today: absolution, or the fine art of hoping problems wander off on their own.

What Do You Design?

Gen AI in Software Development, Not Coding 1

Tacit Knowledge

It Will Do Whatever Fool Thing You Ask It

It Will Do Whatever Fool Thing You Ask It

Agents are Hard

What Is a Developer? Part 2 - Test Boundaries

The Bionic Developer

What Is a Developer?

Retro Computing and Stewardship

London and Chicago Style TDD, A Design Perspective

What AI Tells Us About Ourselves

Structured Semantics

The Bionic Coach

Systemic Oppression and Machine Learning

A Quick Update - Trying to Dig Myself Out

Leveraging Iteration and Decoupling

Diversion #1: Agents

Diversion #2: LangChain & Pydantic

Complex Example

Thinking over Typing 2.0

Simplistic Example

Is Machine Learning a Threat to You?

There Is No Agility Without Technical Agility

I care not for this agile coach, bring me another!

Splitting Product Design and Implementation

Cross-Functional Agile Teams

The Work

The Work

Agile Enough

The Spaces Between

Making a Small Pencil Tray

Functional Decomposition and Customer Centricity

Diversity is Critical

The Cloud Native Virus

Group Kata Facilitation

Pair Chain Exercise

How To Build Good Software vs. Learn a Technology

The Thing You Need On Your Radar

Four Things Impeding Your Software Teams

Coding Is Social

A Nurturing Culture

Lenses

What is an Agile Technical Practices Coach?

Who are you?

Oppression in Machine Learning

Exploring Farming Metaphors for Software Development

Coaching for “likability”

Revolution

The Worthiness of the Fizzbuzz Kata

A Leaf on the Wind

What is a Hacker?

AI Programmers Will Take Our Jobs!

The Long Road

In Memory of Nancy Diamond

Act on Fact, Not Conjecture

Testing Against The Edges

2016 Big Steps Forward

The Fable

Why Kata?

Retrospective on 2015

Lightweight Accountability

Avoid Technical Debt with Planning

Technical Debt and The Big Rewrite

Wicked Code

Signed and Encrypted Email on OSX and iOS

Lies, Damned Lies, and Code Comments

Why a Worker's Cooperative

7 Simple Steps to a Successful Software Project

The Chaos in the Matryoshka

Changes

Decisions In The Dark

Cannibal: A simple Object permissions framework for Ruby

Software Development Is About People

An American iPad in Canada

iPhone Applications and Intermediate Tooling Considered Harmful

Apple: Thank You for Leaving Flash Out

Java EE 6 Mess, Continued

Testable Java EE 6

Java EE 6, Untestable?

Using Java EE 6, Without Java

Java EE 6 and Netbeans 6.8 Released

CRTC, Hear Our Plea

Securing Data on a USB Drive on Mac OS X

Agility, Software Development Estimates and Buying a House

Continuous Improvement: Tools or Process?

The Concept of Millitasks

What You Want vs. What You Need

Outsource? Yes! Offshore? NO!!!

Doctors are people too

My Ongoing Battle with Pseudomyxoma Peritonei (PMP)

Doing Waterfall Badly

SSL-Explorer killed by Barracuda ?! Forked as Adito ?!

Software For Business Is Not Just A Tagline

Today

Off To Find My Shoes

Taking a stand, saying NO to software development quotes

Integrity, Transparency and Trust

Dissapointed In Dell

VMware ESXi At Long Last

iPhone Blogging 2.0

iPhone Blogging

Wishing for the days of XML Deploy-time configuration??!!

Performance Profiling with JBoss Seam - Part 2

Performance Profiling with JBoss Seam - Part 1

Living with Maven2

Keeping A Lid On It

Learn Your Tools, Learn Others Too

Software Development Estimates

Database Queries - Programmatic Criteria vs. EQL?

Why, oh why, Wordpress, have you made it so difficult to publish code snippets.

Wow, this isn't the Web Services I used to know...

Like we needed more proof that DRM is a BAD IDEA

Property Binding in UIs

Am I a whiny little bitch because I want Java6 on my Mac?

Disappointed.

Sanitizing your input considered harmful.

Simplicity Lessons from EasyBrandingTools.com

Lies, Damned Lies, and Software Development Estimates

Keep It Simple, (Not So) Stupid

End User Experience - Why, Microsoft, why?

Using 'Open Source' Commercially

Unit Testing: It's About Confidence

Speak Clearly from Experience

Software Developers

Communicate on Their Terms

Hello again!

How to answer 'What's wrong with PHP?'