Dragan Stepanović

I'd vote for "Repeat three times and you keep it".

I literally had a conversation with one of the managers I work with where I asked him point blank: “if I brought you results that changed every time you looked at them, how would you feel about that?”

He said he wouldn’t trust them.

Then I asked him why he was trusting ChatGPT results.

Tests tell us our software can work.
Observability tells us our software IS working.

To understand that I’m there to listen and learn from all the signals that can teach me — if I’m ready to be taught.

To attune to the world around me, as I’m a part of it — not to forcefully impose my fantasy of what I think it should be.

3/3

To listen to other humans and their experiences as we create and shape together.
To sense what a sustainable pace feels like.
To move as an interconnected and interdependent whole.

2/3

It’s been a long time coming, but what XP ultimately taught me was to listen more — and to listen carefully.

To listen to what the code is trying to tell me, and follow the path it wants to evolve toward.
To listen to tests, and how they can help me mold the design.

1/3

🔬 Real Browser Behavior Modern browsers like Chrome, Firefox, Safari, and Edge all follow this exact logic — because they conform to the official CSS standard. If you try this: html Copy Edit <style> #box { color: red; } /* (1, 0, 0) */ .a.b.c.d.e.f.g.h.i.j.k { color: blue; } /* (0, 11, 0) */ </style> <div id="box" class="a b c d e f g h i j k">Hello</div> The result will be blue text, because the rule with 11 classes has higher specificity.

A problem with modern LLMs is that they are so convincingly anthropomorphic now that when they produce wildly incorrect responses like this one, my reaction isn't "oh, I found a bug", it's "you're a lying sack of shit!"... and that level of emotional reaction to a piece of code really isn't healthy.

Whatever your problem might be, I heard you just need to add more rules in the .cursorrules file.

I heard you just need to add more rules to the rules file.

We don't do politics here, but politics does us.

No techno-fix.

It's like Fox News!

If you reduce the cost of building features I guess what you'll get is more features.

I hope that's what customers actually needed.

Using AI to create big batches of code faster instead of using it for sooner delivery of thin slices - which, by the way, I don't think was ever a constraint in product development, but anyway.

That's what you get when slapping a tool on a wrong underlying mindset.

Doing wrong things faster.

On the Benefits of Trunk-based Development

Published: On the benefits of trunk-based development

Whenever we suggest trunk-based development, we are called unprofessional and belittled. How can we possibly produce quality? But in all honesty, for trunk-based development teams, quality is the highest concern.

thinkinglabs.io/articles/202...

The lack of (explicit) prioritization doesn't mean that prioritization is not taking place.

When there's too much WIP in the system, prioritization is inherent, just that it's implicit and someone that has way less context than the one that should prioritize ends up doing it.

"Post-Doom teaches that, ironically, it is the very urge to cling to hope and the faith in progress and tech that is driving us faster and faster toward our own annihilation. When we refuse to acknowledge natural limits, then we end up hastening the very outcome that we want to avoid." John Halstead

"but it's simpler!"

If I had a € for every time I saw “if err != nil” in a Golang codebase, I'd be a trillionaire.

We ran a randomized controlled trial to see how much AI coding tools speed up experienced open-source developers.

The results surprised us: Developers thought they were 20% faster with AI tools, but they were actually 19% slower when they had access to AI than when they didn't.

The irony is that dev teams who work in ways that appear to get the best results from LLMs - small steps, prompting with tests, rapid feedback loops, continuous testing, code review and integration, and good separation of concerns - have little need for the tools.

Neoliberal systems thinker is an oxymoron.

Oh, I have another one! Available for pre-order!

2.Code Churn This metric measures the code added, modified, or deleted over time. It indicates how stable or volatile the codebase is and how much effort teams spend on maintaining or changing the code. A high code churn means the code is constantly changing, which may imply that the code is poorly designed, the requirements are unclear, or the developers are not following best practices

Or... Or... the team has mechanisms and practices in place to keep the cost of change low, so they are able to change code and adapt design just-in-time (emergent design), rather than trying to predict it when they have the least information about the code (design up front crystal ball gazing).

New t-shirt in my store.

Change in Western Europe Summer Temperatures Data from Berkeley Earth. Anomalies relative to the 1850-1900 The curve starts exponential rise in the 1980ies

I'll just leave this here

Thriving in a dysfunctional environment was never an indication of progress.

Not new, but the TLDR; of this "You probably don't need MediatR" article is perfect 🙂

arialdomartini.github.io/mediatr

"Anyone who thinks you can have infinite growth in a finite environment is either a madman — or an economist."
- Kenneth Boulding