12 February 2026

34 links · Thursday

Introducing GPT‑5.3‑Codex‑Spark

Introducing GPT-5.3-Codex-Spark—our first real-time coding model. 15x faster generation, 128k context, now in research preview for ChatGPT Pro users.

Major European Payment Processor Can't Send Email to Google Workspace Users

Viva.com, one of Europe's largest payment processors, sends verification emails without a Message-ID header — a recommendation of RFC 5322 since 2008. Google Workspace rejects them outright. Their support team's response to my detailed bug report: your account…

Turning YouTube Into Cloud Storage

How I made a YouTube file media storage using C++ and a few libraries. You can view my repository here: https://github.com/PulseBeat02/yt-media-storage Here is the Wirehair O(N) fountain codec repository: https://github.com/catid/wirehair For those who want …

Markdown.exe

We have turned markdown into an executable file. Agent use skills, and threat actors are using it to spread malicious scripts all over the web. md is exe now

Exploring Chess Positions and Counts

Introduction In our earlier article “How Many Chess Games are Possible?” we derived that the number of typical short chess games is around 10151. This turned out to be much larger than …

On Seeking Order in Chaos

Generalist software developer writing about scalable infrastructure, fullstack development and DevOps practices.

But What To Do About It?

The future is here. I'm deeply uncertain what to do about it. I first had a passive inkling AI might be much bigger than chatbots 2 years ago in Alaska. Like buying my first bitcoins in 2013, I just paid a little extra attention when smart people around me st…

Dorodango

I used to write more

The Problem With LLMs

Premise Six months ago, a friend of mine, with whom I work on the nonprofit Pariyatti mobile app, sent me this blog post by Vijay Khanna: From Idea to App in 7 Hours. By now, this is a fairly common zero-to-one LLM coding story. (LLM is short for Large Languag…