Week 47, 2024 - Talk Talk

Week 47, 2024 - Talk Talk
Budapest, 6:35 AM, 22nd November, 2024.

This morning I woke up to a beautiful, snow-covered city. Winter is my least favorite season, I often find the dark, cold days bringing down my mood. Still, I try to focus on the good sides of it, and fresh snow is definitely one of the few ones.

I spent a big part of Tuesday taking apart an old sofa, moving its parts to the garage, and then assembling a new one. The physical work, stepping away from the computer, and then enjoying the result of my efforts was a nice reward. (And also a proof that the IKEA Effect is working - I love our new furniture).

📋 What I learned this week

I gave a talk to managers of a tech startup this week, explaining the importance of building trust and motivating people during challenging times. Based on the questions and comments, the high-pressure era we’re in is tangible. Delivery pressure, fast-changing priorities, reorganizations, engagement and retention issues, burnout. I believe that these periods come and go, the pendulum swings to one extreme before correcting itself again. Building resilience to be able to support people and nurture human relationships is the way to get through high-pressure eras like the one we’re entering after many years of peaceful growth.

It’s been a long time since I tinkered with my home lab, so one evening I tried AdGuard Home as a replacement for my local DNS server. The results were good, and since I have two physical nodes now (and there’s a great project that syncs settings between them), I could finally serve both primary and secondary DNS servers in my router for the whole family. I used to have a more nuanced view on ad blocking when I worked in online media, but the impact of bloated AdTech platforms on browsing performance, together with the parental filtering features converted me to a believer in the right to choose what I want to look at, and what I don’t.

🤔 Articles that made me think

Lessons from an Obscure Bug

These investigations are always fun to read, almost like a crime novel. But I especially love the follow-up I linked in the subtitle. The engineering seniority of the author is clear from his explanations of past decisions, focusing on learning, and making pragmatic decisions in mitigating the issue first (a quick and dirty bug fix stopped the pain for users) and then refactoring the offending area.

It’s awesome to see the main goal of all their refactoring decisions: to help future developers working with the code. In teams, especially bigger organizations, I believe one of the main roles of developers is to unblock their peers. It informs every decision: prioritizing the review of PRs, building self-serve systems and documentation, and yes, optimizing refactoring for future readability.

The end of the API era?

Strava announced draconian changes in their API usage terms. DC Rainmaker is doing a great job explaining all the details and aspects of why and how these changes can be brutal for a blooming economy of third-party apps, from Intervals.icu to VeloViewer. Arguably, any decent app built on the platform will violate these new terms and has 30 days to remediate. I wonder how my tiling hobby will survive this - if anything, at least the the progress updates will probably have to be more manual in the future.

I remember when Twitter launched and quickly conquered the world in the second part of the 2000s, and one of the key aspects of its success was an open and free API. Taking an API-first approach to product development was a great strategy to increase active users and similar numbers showing the popularity of a product, critical KPIs in an era of exponential growth. Let your users build upon your product, let them find new ways to use it and abuse it — and implement or acquire the most popular ones. The cost of operating an API is irrelevant when you’re swimming in free money and fighting for a bigger slice of the cake. A nice milestone to signal the end of this era was when the once API-pioneer Twitter cut down drastically their API usage terms, And now, Strava, an app that is (was?) the de-facto sports activity platform for millions is following this path.

High-pressure eras force you to minimize the number of bets you’re taking, cut off all slack, and focus on a clear, monetizable strategy. It often requires pivots away from earlier strategies, and probably this is what we're seeing in Strava's case too.

🎯 What I want to try next week

Responding to the invitation to fill the hole of a missing speaker, next week I’ll be presenting at the Craft Mini Conference in Budapest about staying relevant as a developer in the current tech era. The opportunity is both exciting and terrifying: it's been a long time since I spoke in front of a big live crowd. I only have a week to prepare, so I plan to spend most of it to make sure there’s good value in return for the audience’s time. If you’re in Hungary on the 28th, they might still have a few tickets left.

Two things going live next Tuesday that I worked on, a new episode of our podcast with Jeremy Brown, The Retrospective, where we’ll talk about Tech Debt, and a guest post to the Besides Code newsletter about one of my favourite topics: the transition from a developer to a manager. You can sign up for both to get notified right when they launch!

That’s it for today, assemble something you'll love this weekend,

Péter

Subscribe to my newsletter

I write about engineering leadership topics.
Sign up to receive new articles.
jamie@example.com
Subscribe