Issue 27 - The Orthogonality Factor

A summer writing

I recently engaged with smart experienced craftsmen, both at work and outside. The wisdom and knowledge coming from such folks often has a shocking effect on me. It often needs one word, one observation, and my brain is flipped 🤯.

Those last two month it has been about: orthogonality. Two people, in uncorrelated way, told to me about orthogonality. To quote one of the two (here Lloyd Tabb) :

If you have been in the computer business for any length of time, you have encountered software that is NOT orthogonal; software that is arbitrarily complex, full of special cases where things don’t work.

As it’s really relevant in software, my mind has wandered how this concept also apply in usual society issues and systems broadly speaking…

But I won’t spoil myself: go to the 📰 Blog Post part of this post where I link a whole article summarizing my thinking :)

---

📡 Expected Contents

The next crisis : technical debt

A bit of alarmist headline, but I 100% agree with this blog post: technical debt has taken over the economy.

From financial services, transports to manufacturing, important parts of our globalized economy depend on services designed and created more than 20 years ago.

Hopefully, there haven't been big changes and those systems were well designed enough to support our current context.
Still, it seems some of them would need a complete migration at some point. Like our databases.

Code changes take weeks, design changes take years.

---

Malloy, again

I think sometimes I could be a sales guy 😅.
I'm completely hyped by Malloy recently. The (google) team behind is very friendly over their slack and while Malloy is an experimental langage, it already conviced me with its ease of read and integration with DuckDB.

---

For the love the game

What if we design data experience like games? Instead of showing mindless numbers and only filters over filters we could draw a real user experience, with challenges, targets and levels to reach.

That's the whole idea of the last dbt newsletter written by Windy. Very nice read it really resonate with my current feeling of our experience with data in general...
Dashboards are for years the data driven end game while they should be the game menu.

We should seek experience. Not for the sake of it; but because this is the way we draw our intuition.

---

Data is not a Microservice

Maybe one my favorite read of the summer. Chad explains how software engineering can't solve data's problem.

I often say that data is 5 or 10 years lagging behind software. Data mesh, CI/CD, tests, code versioning, etc.
All the major changes which happened in the data industry come from software mindset.

Still, I agree with Chad: data has something of its own.

The purpose of a microservice is to power an aspect of some customer experience. Its primary function is operational. The purpose of data is decision-making. Its primary function is TRUTH. How that truth is used can be operational (like an ML model) or analytical (answering some interesting question).

I let you read the whole piece, but we should not think that all will be easy and just follow blindly the software industry. Like always, we should inspire our practices from different fields, but we should also create new ones.

---

📰 The Blog Post

Scaling Smarts: The Orthogonality Factor

This is a kind of “summer” article coming here: more a personal thinking put on paper ✍️

---

🎨 Beyond The Bracket

That’s the kind of unexpected opportunity arising from passion…
Last weeks I was honored to have one of my old data-visualisation exposed in Palais des Congrès in Paris 😊

Curated by the awesome Marthe Viallet, my data-viz has been selected by Christopher C., Head of Performance at FFF - Fédération Française de Football 🇫🇷⚽

All this side-project started on Twitter 5 years ago¹... seeing this printed on a wall for an exhibition during a major sport & science event (ECSS2023) is unbelievable to me 🙏

Beyond showing up, it remembered me how much passion can bring. At the time I was not looking for anything: just making fun out of doing data-viz and learn tech skills.

This is a feeling hard to find as we're all nudged by the "time is money" and activities driven by some outputs.

While we should have goals and dreams, we should first love the process (again) and let the passion vibrate into ourselves.
Sometimes it won't go anywhere, but at least you really lived through it. This is priceless.

---

I hope you had nice summer vibes ✨.

I felt that round of wisdom coming from slow pacing my activities and consumption of everything. Focusing on sport, books and music.

Sometimes it’s good to slowdown and breath. Especially during summer. Where nights shine and the sun radiates in its dullness ☀️.

Now it’s back to school - many things in the pipe on my side. Let’s go back to work 💪

Parisians during summer

1. If you wonder how I made those charts, you can check this old GitHub repository where I used to do all my data-viz back when I was doing football analytics. Also this article, describing the motivation and explanations behind those plots.