Code reviews can't keep up varoa.net/2026/02/22/c...

Just posted varoa.net/2026/02/07/a...

Optimizing software for agent-legibility is quickly becoming a hard requirement... Screenshot from the last "Joy & Curiosity" issue by @thorstenball.com More on identifiers: varoa.net/2024/05/01/i...

“[…] writing, which is the technology of the rational mind”

What Doesn’t Change

One of the interesting parts about software engineering is that each domain you work in requires re-asessing every pattern and best practise that seemed set in stone in other domains.

I have Copilot (+ Claude) stuck in an infinite loop of refactors with small "improvements" to its own implementation of a relatively simple prompt. Feature parity with human devs is one step closer.

Listening to The Sonics' Hitch Hike, noticed that The Velvet Underground reused the riff in There She Goes, the Sonics were covering Marvin Gaye. www.youtube.com/watch?v=6-US... www.youtube.com/watch?v=KwyL... www.youtube.com/watch?v=QbW3...

Software engineer -> Software operator www.complexsystemspodcast.com/episodes/the...

Warming up to this reading list enforced backupcy thing, thanks Pocket

It takes you about the same amount of time to read this post as the amount of downtime you’d be allowed per week with 99.999% availability

“One does not replace a blown fuse until one knows just why it has blown.”

There are still 1980’s style sysadmin orgs that have survived by hiding behind the fancy labels of DevOps / SRE / Platform engineering

Been inspecting some pmap outputs these days, and could not find any decent visualization tools. Here is a quick one (nods to @simonwillison.net toolbox). varoa.net/tools/pmap-v...

New small tool coding workflow: open vscode, agent mode, ask for tool, go do $chore

So em dash is the new delve?

The more we lean into vibe-based coding, the clearer it becomes that we’ll still need good operators to keep everything sane and under control

Surfacing this from a thread.

Queuing theory: probably the hardest topic to confine to work. brooker.co.za/blog/2025/03...

Everything associated to the “software craftsmanship” mystique is up for surprises

When I wrote about "AI will choke delivery pipelines", no-Fri-deploys was one of the practises I mentally labelled as AI-unfriendly. Interrupting the flow of code on Fri creates a glut for Mon. And bundling changes tangles latent defects, making troubleshooting harder. varoa.net/2025/04/07/a...

When bundling $thing2 into $thing1 looks like a good opportunity, that's the signal for NOT doing it. If they looked like bad opportunities overengineering & overscoping would never happen.

Anybody want to come work on Tailscale's internal dev tooling/productivity, CI, bots, dashboards, etc? Pretty please? 😅

I mention the DORA report in the piece I published today. The report has a page about AI impacting sw delivery negatively. The recommendations (screenshot) about smaller changelists and robust testing make sense, but I'm somewhat skeptical that they are viable ->

If you want a TL;DR of the post it’s probably this one!

"AI-generated code will choke delivery pipelines", just published. varoa.net/2025/04/07/a...

The type of content I expect from my timeline on a fine Monday like this

Useful != worth doing

Reproducible builds ftw lwn.net/Articles/101...

Also a good way to treat clusters as cattle

Me he abierto un substack. Para hablar de libros. Pero he acabado montándome una peli con Truffaut, Policán y Penélope. Pido perdón. open.substack.com/pub/pepogarc...

A lot of people don’t realize that putting all build / logic in GHA definitions they essentially turns GHA into an operating system for their dev environment. Which is, of course, a bad idea.

While we are at it. Even though people act otherwise, semver declares intent but offers no guarantee. Hopeful versioning might have been a better name

Less reminded than learnt from the first time. Can't count the number of discussions I've had about using git tags as source of truth for release procedures and building artefacts. (From the NH thread about the tj-actions/changed-files vulnerability) news.ycombinator.com/item?id=4337...

The article is about turning around platform teams but some points apply to any struggling eng team https://buff.ly/3QlEK7q

Same with movies (with letterboxed) I miss the old @lastfm.bsky.social recommendations

Adding to my reading list. Moving data across networks is a major challenge with AI (both due to the amount of data, and the different needs of each workload, e.g. ingestion, training, inference, etc.)

Great read dasl.cc/2025/01/01/d...

Cross posting bc. @rakyll.org didn't and it's massively important. 2 comments 1) We have decades of mgmt structures evolved based on a playing field where large scale coordination and alignment were essential. They are now obsolete. Some will take years to unwind them and will get overrun ->

// TODO: is underrated as issue tracker

Spring boot configuration, that's the true AGI test

The "AI Architect" point is actually the least interesting topic in the podcast with Bret Taylor, thinking AI beyond autocomplete, and lots of good stuff. podcasts.apple.com/us/podcast/t...

"The LLM Curve of Impact on Software Engineers" hits the nail on the head regarding the usefulness of LLMs for coding. As the maintainer of a medium-sized and fairly custom codebase that I wrote myself, I very much fall within the trough labelled as Senior here.

Looking forward to “The Mythical AI Coding Agent”

A good read on how DynamoDB has evolved over the past decade. DynamoDB was originally launched in 2012, and this retrospective paper was published in 2022. www.usenix.org/conference/a... Some takeaways after reading it 🧵

Really not a fan of the YouTube delivery style, but the content is worth it.

Happens to every label and it’s the reason why it’s better to focus on individual principles with direct application in real situations, and avoid engaging with the labels.

Pretty close to my own experience so far on small projects, prototypes and such.

The biggest winner of the AI race will be distributed systems people. Everything is converging onto a distributed network of stuff and it is only accelerating in the last two years.

Experimenting with mutt + @simonwillison.net 's llm for mailing lists: macro index,pager ,a "| llm -m gemini-pro 'Summarize and highlight key points, original ideas and actionable content. Remove invitations to subscribe to mailing lists. Format for max line width of 80 chars' | less"