Programming

Do you remember the recent post that self-assessing any effect is hard? Here is a comparison between self-assesment and measurements.

Surprisingly, we find that when developers use AI tools, they take 19% longer than without—AI makes them slower.

Before starting tasks, developers forecast that allowing AI will reduce completion time by 24%. After completing the study, developers estimate that allowing AI reduced completion time by 20%. Surprisingly, we find that allowing AI actually increases completion time by 19%

N = 16

Extra.

LOL in a lot of ways.

2025 is gonna be interesting.

Hey everyone — I'm a final-year student, and I’ve been wondering this a lot lately. We always hear that “you need a good project to land a job”, but most students I know either copy from GitHub, get stuck, or just... give up. We’re doing a small open survey to understand this from both sides — students and educators. If you've ever: Built or struggled with a final-year project

Helped someone else do it (educator/mentor)

Wanted to sell or learn from real-world projects

We’d love to hear your honest experience. 🙏 It’s just 2–3 mins, totally anonymous. 📄 Survey Link – for students & educators

We’ll be using the insights to create open resources and maybe a system that actually helps. Thanks in advance if you participate — or drop a comment about your experience.

The population (especially the younger generation, who never seen a different kind of technology at all) is being conditioned by the tech industry to accept that software should behave like an unreliable, manipulative human rather than a precise, predictable machine. They're learning that you can't simply tell a computer "I'm not interested" and expect it to respect that choice. Instead, you must engage in a perpetual dance of "not now, please" - only to face the same prompts again and again.

This article is divided into two chapters.

• The first one, starting with section “Introduction to naming in programming” presents a review of scientific literature present on the topic. That section will deepen your understanding of the current body of knowledge on naming things.
• The second chapter, starting with section “Guidelines for naming conventions in programming” presents actionable recommendations to improve your skills in choosing thoughtful class, function or variable names. If you’re looking for tips, go there.

There are only two hard things in Computer Science: cache invalidation and naming things.

I have seen Geany around and am pretty intrigued by the update. I primarily use Kate but would love to hear folks' thoughts on Geany and if anyone recommends it.

I have heard about Elninki's and Harvard's programs, can I join them now or is there a specific date?

Hey everyone, I’ve been working on something in my free time — a wrapper that lets you use open AI models (like Mistral, LLaMA, etc.) in a clean interface. The longer-term idea is to make it easier to run larger models — 70B+, fine-tuned variants, or your own — without needing your own GPU cluster. Still very much a side project — I’m building this for fun, learning as I go, and curious if it’s useful to anyone else. Link’s here if you want to try it: https://umbraai.xyz/

I wanted a dead-simple way to share text, links, or code without creating accounts or dealing with messy UIs. So I built Zync(https://zyncshare.vercel.app/) — paste, share, done.

• Share plain text, links, or code instantly
• Replies work without login
• Everything auto-expires in minutes or hours
• No tracking, no ads, no clutter.

Would love to hear what you think. Is it something you’d actually use?

This is the name of a concept to combine source code and highly structured human-readable documentation, such that the main artefact is the documentation, and the secondary artefact the full source code extracted from it. Any change to the source means changing the explanatory documentation of the goals and reasoning at the same time. It was invented by no other than Don Knuth, and he, together with his collaborators, wrote the TeX program in it, to show how it is useful in practice.

I myself have used the technique in various occasions at work, for example when taking on an important but almost unreadable piece of code I inherited at work, to preparing sources for a complex algorithm for handover when I was leaving another larger project, and also in leisure coding.

The tool which is most widely used for this approach today is Emacs org-mode. It has the advantage that it works for every programming language, and you will likely find full examples for all languages you use, but many other tools (for example for vim) exist.

Andrew Zig Roadmap 2026.

It's a 2h talk with code / demo. A lot about the new Io interface.