I’ve had the pleasure in my career of sitting in tons of different board rooms talking with executives and getting a rundown on different charts and graphs they care most deeply about. As expected, “number chart go up” is always the desired outcome. Number chart go up is good because it makes other number charts go up, specifically stock ticker number.

Instead, I act like a sponge. I listen to the stream of day-to-day noise, absorb the problems people are having and let them sit in the back of my mind. Over time, some fade away while connections begin to appear between others that initially seemed unrelated. Eventually, I start to see what’s really slowing people down and what my team or I can do about it.

A guide to preventing Postgres from toppling over.

You read that right

Every few months or so, somebody asks on social media why a particular DVLA digital service is turned off over night. Rather than answering it every time, I’ve decided to write this post, so I can point people at it in future. It’s a great case study to show why making government services digitally native can be quite complicated.

We built a social network on SQLite, put it behind a Node API, and hit it until something broke. One file, one process, 315 million requests a day on the heaviest endpoint. Here are the numbers, the config, and the honest limits.

La canicule en France entraîne toujours un débat sur la climatisation. Si elle est une commodité que personne ne questionne dans des tas de coins du globe, elle jouit d'une mauvaise réputation historique dans l'Hexagone, basée sur des croyances qui cachent les éléments factuels permettant d'élever les discussions.

Un beau matin, ma voiture n’a pas démarré : rien ne répondait. Ni la clé à distance, ni le bouton sur la porte. Une fois dedans avec la clé d’urgence (une serrure derrière la poignée de porte), impossible de démarrer. J’avais déjà connu ça sur m…

From a coworker at Mural

Each claim has to pass the IRS' 4-part test:

1. Permitted Purpose: The activity must aim (not necessarily succeed) to create or improve functionality, performance, reliability, or quality of a component of the business.
2. Elimination of Uncertainty: At the beginning of the project, there must be technical uncertainty regarding the capability of development, the methodology, or the optimal design.
3. Process of Experimentation: The activities must involve evaluating alternatives—such as modeling, simulation, or systematic trial and error—to resolve the identified technical uncertainties.
4. Technological in Nature: The research process must fundamentally rely on the principles of hard sciences, such as physical or biological sciences, engineering, or computer science.

#1 and #4 are pretty much always a given, and then ICs get called in to audit-proof #2 and #3. The painful part is that our record-keeping is much better at answering questions about outcomes, not the experimental process or all the exploration that never ended up getting shipped.

For roughly the last ten years, a meaningful percentage of my working hours have been spent thinking about observability. If you're not familiar with the term, "observability" is what we call it now that "monitoring" doesn't sound expensive enough. The actual work is unglamorous in that you collect

That’s something and that’s not great feeling for the future

Stop naming booleans flag or done. Learn the four prefixes (is, has, can, should) that eliminate ambiguity, prevent bugs, and make code reviews faster.

A blog about software craft and quality

We've all worked with a rockstar developer. They joined the team years ago, full of energy. They had great ideas about new tech, new paradigms, new...