Developer Diary: February 2026 - From Email Features to AI Skepticism

Shipping Features and Learning New Things

This February marked a milestone for me at Ghost - shipping my first feature: Inbox Links. When members enter their email to log in or sign up, we now show a button that takes them straight to their inbox. Beyond the satisfaction of shipping a useful feature, I got to dive deep into some fascinating technical territory.

Learning about MX records and RFC-compliant email address parsing was particularly enlightening. Email infrastructure is one of those things that seems simple on the surface but reveals incredible complexity once you start digging in. The source code for the main logic is available here.

Unexpected Technical Discoveries

One of the most surprising things I learned this month was that gzip streams encode which operating system did the compression. I had no idea this metadata existed! This discovery led me to build a little tool called "gzpeek" to inspect this metadata and more. It's these kinds of small, unexpected technical details that make programming endlessly fascinating.

Celebrating Gaming History

February also marked the 40th anniversary of the original Legend of Zelda - a game that fundamentally shaped the action-adventure genre. Inspired by this milestone, I created a calendar feed for other game anniversaries. It's amazing how these retro games continue to influence modern game design. Just yesterday, Oracle of Ages and Oracle of Seasons turned 25 - time flies!

I also wrote a few articles for Zelda Dungeon as usual, exploring various aspects of the series and its impact on gaming culture.

Cool Links from Online

I came across some fascinating content this month that I wanted to share:

• An incredible stunt: a game written entirely in HTML and CSS. "No JavaScript or server-side code is used." The creativity and technical skill required for this is mind-blowing.

• The best description of LLMs I've seen so far: "When you enter text into [ChatGPT], you're asking 'What would a response to this sound like?'" This captures something essential about how these models actually work - they're pattern matchers, not thinking entities.

• A thought-provoking article titled "AI skepticism is a quiet career killer" suggests that being bullish on AI seems better for your programming career. This raises interesting questions about the relationship between technical skepticism and career advancement.

Global Tech Politics

I've been following the growing tension around US technology firms' global dominance. "Countries are growing uneasy about their dependence on US technology firms," and the US is trying to stop this shift. I believe we deserve to lose this battle. The lack of European billion-dollar technology companies leads people to forget the technology invented here that instead embraced openness: the web, Linux, Raspberry Pi, Open StreetMap, the Fediverse.

Digital Privacy Awareness

I appreciated an article about someone who decided to pick up a retro video game instead of doom scrolling. It's a simple choice that speaks to larger questions about how we spend our time and attention.

Another piece, "What Your Bluetooth Devices Reveal About You," made me think differently about wireless security. "This isn't about paranoia. It's about understanding the trade-offs we make when we leave wireless radios enabled on our devices. For some use cases, Bluetooth is essential. For others, it's just convenience. Being aware of what you're exposing is the first step to making informed decisions about which category your devices fall into."

Security and Ethics

I found a great breakdown explaining how Content Security Policy actually works and, more importantly, where people screw it up. CSP is one of those security features that's powerful but often misunderstood or misconfigured.

On the environmental front, research showed that "the research [...] did not find a single example where popular tools such as Google's Gemini or Microsoft's Copilot were leading to a 'material, verifiable, and substantial' reduction in planet-heating emissions." This challenges some of the narratives around AI's environmental benefits.

The Problem of Openwashing

I first encountered the term "openwashing" this month, which describes the process of using the word "open" as marketing. OpenAI and Android are examples of this - companies that use "open" in their branding while maintaining significant control over their ecosystems. The article "Acting ethically in an imperfect world" explores this concept further.

Looking Forward

February was a month of shipping features, making unexpected discoveries, and reflecting on the broader tech landscape. From the technical details of email infrastructure to the philosophical questions about AI and openness, there's always something new to learn and consider.

Hope you had a good February too.