Coding matters: Please, thank you, and good spelling
Jacqui Coosner
2026-04-17
Do you thank your AI? Does politeness get you better results? A quick look at the confusing research, and what 3 LLMs claim.
Coding Matters
Jakarta EE versus Spring
Lewis Coosner
2026-04-09
Jakarta EE and Spring Boot are the two de facto frameworks when developing enterprise Java applications. They are both very powerful frameworks. However, they have different design philosophies and cater to different development needs.
Jakarta EE 12
Lewis Coosner
2026-04-02
The Jakarta EE 12 release cycle is currently one of the more interesting developments in the Java ecosystem.
State of the Java Ecosystem: March 2026 Update
Lewis Coosner
2026-03-27
March 2026 update on the state of the Java ecosystem. It is faster and leaner than 3 years ago.
Coding matters: Bills, Bots, and Bad Systems
Jacqui Coosner
2026-03-27
The Annoyance Economy is real. It takes the form of unhelpful processes, AI bots, and bad systems. You’re not paranoid – some companies are out to get you.
Coding matters: Lost in translation
Jacqui Coosner
2026-03-20
Ever heard the phrase: *“If you can’t beat them with brains, baffle them with blather”*? Here are some ways to explain project problems in Boardroom Babble or Tiktok Talk.
Coding matters: Debugging with a hammer
Jacqui Coosner
2026-03-13
Rage rooms are all the rage, even in South Africa. They could offer packages with a more technical flavour. This could be a way to vent your frustration with everything tech, including AI.
Coding matters: Teaspoons, chocolate, and AI
Jacqui Coosner
2026-03-06
A light look at the mystery of disappearing teaspoons in corporate offices. And AI’s explanation for it.
Coding matters: Fast bots and frazzled humans
Jacqui Coosner
2026-02-27
AI is getting faster, but people aren’t. We deal with warnings about AI taking over tasks, and the real fatigue of switching between too many jobs. And then AI still hallucinates.
Coding matters: Scary software stories
Jacqui Coosner
2026-02-20
Big software failures rarely begin with big mistakes.
More often, they start with something small: a data conversion, a timing issue, a line of code that looked perfectly reasonable — until real-world conditions proved otherwise.