Vincent Behar

As we test Kubernetes clusters beyond the supported 5,000+ nodes limit at LinkedIn, we're finding out 'List API' calls are a bottleneck in cluster reliability. So I'm publishing a deep dive about why this is a problem and what you can do about them. ahmet.im/blog/kuberne...

for function calling, there's still a big diff between proprietary models and open-weight models...

Prompt engineering is dead, long live Context engineering!

Well, maybe something like 1 year? ;-)

Sneak pick:
- Kubernetes: Origins
- The Crew: One UX to Drive Them All
- Prince of Platforms: The Sands of Innovation

#Kubernetes #SunnyTech2025

GitHub - charmbracelet/fang: The CLI starter kit The CLI starter kit. Contribute to charmbracelet/fang development by creating an account on GitHub.

Charm published a new Go CLI starter kit: Fang (experimental)

github.com/charmbracele...

Features
- Fancy output
- Fancy errors
- Automatic --version
- Manpages
- Completions
- Themeable
- Silent usage output

Need to test 😇

Blog: Changes to Kubernetes Slack-

Nice! The interesting part is github.com/apple/contai...

TL;DR it uses a lightweight VM for each container

Introducing container use for agents.

Go from babysitting one agent at a time to enabling many agents to work safely and independently with your preferred stack.

github.com/dagger/conta...

For my simple use case (read it later / offline) instapaper is really good so far

Instapaper A simple tool for saving web pages to read later on your iPhone, iPad, Android, computer, or Kindle.

I just tried www.instapaper.com to replace Pocket and found out I already have an account.

Indeed, there was 1 item in my account from 14 years ago!

> Automatic Flushing: The Rails 3.1 Plan

Goodbye pocket... support.mozilla.org/kb/future-of...

Now I need to find an alternative...

When the LLM doesn't know what LLM means...

In an AI-generated code review, it suggested the following comment:

> "DAGGER_ALLOW_LLM=all", // Enables Long-term Memory (LLM) capabilities for Dagger processing

The "dev & contribution" section is great: deepwiki.com/dailymotion-...

Even when there's decent documentation for an open-source project, this part is usually the least documented. This is really useful.

DeepWiki is pretty cool. I've tried it on some on my own projects, such as octopilot, and the generated schemas are really nice.

For example Octopilot has a complex logic to create/update github pull-requests with 3 different strategies, and deepwiki.com/dailymotion-... makes it easy to understand!

We just released @updatecli.io version v0.100.0! While hitting version 100 feels like a milestone, the truth is even more impressive: we’ve published 206 releases since the beginning of the project, 5 years ago.

www.updatecli.io/blog/updatec...

#devops #continuousdelivery

because it's simple and human-friendly ;-)

Mason: a declarative build tool on top of Dagger | Vincent Behar Mason: a declarative build tool on top of Dagger I wrote something. A build tool. A declarative build tool on top of Dagger. Why? Because I really like Dagger, but as I tried to spread its usage at ...

I wrote a declarative build tool on top of Dagger. Say hi to Mason! whtwnd.com/vbehar.dev/3...

More details at github.com/vbehar/mason

Congrats, you're a front end developer now :-)

What ? You don't do that too? ;-)

Nice! And perfect timing with the buzz around nats ;-)

GitHub - akhenakh/narun: Narun is a lightweight HTTP gateway designed to bridge incoming HTTP requests to backend services communicating over NATS Narun is a lightweight HTTP gateway designed to bridge incoming HTTP requests to backend services communicating over NATS - akhenakh/narun

Releasing a new experiment: Narun is a lightweight gateway (autonomous or on top of Caddy) that bridges incoming HTTP or gRPC requests to backend services built using the NATS Micro framework.

github.com/akhenakh/narun

Let's fix OAuth in MCP Let's not overthink auth in MCP.

Excellent blog post by @aaronpk.com on OAuth for MCP aaronparecki.com/2025/04/03/1... - thanks!

Et alors, tu l'as convertie au scala ? ;-)

Here's the main prompt that runs in a loop, which starts like this: You are operating as and within the Codex CLI, a terminal-based agentic coding assistant built by OpenAI. It wraps OpenAI models to enable natural language interaction with a local codebase. You are expected to be precise, safe, and helpful. You can: - Receive user prompts, project context, and files. - Stream responses and emit function calls (e.g., shell commands, code edits). - Apply patches, run commands, and manage user approvals based on policy. - Work inside a sandboxed, git-backed workspace with rollback support. - Log telemetry so sessions can be replayed or inspected later. - More details on your functionality are available at codex --help The Codex CLI is open-sourced. Don't confuse yourself with the old Codex language model built by OpenAI many moons ago (this is understandably top of mind for you!). Within this context, Codex refers to the open-source agentic coding interface. [...]

Wrote some notes on OpenAI codex, their new open CLI "agent" tool for writing and iterating on code. Since it's open source (Apache 2) the workings are available, including an interesting system prompt and a macOS sandbox using the sandbox-exec mechanism simonwillison.net/2025/Apr/16/...

The models are choosing their names themselves... And "sometimes" they get confused ;-)

"Kubernetes and Model Context Protocol discussion" groups.google.com/g/kubernetes...

GitHub - minio/minlz: MinLZ is a LZ77 compressor, focused on realtime data compression MinLZ is a LZ77 compressor, focused on realtime data compression - minio/minlz

I with just released about 2 years of work on improving compression with a fixed encoding LZ77 style compressor for #golang

Our goal was to improve compression by combining and tweaking the best aspects of LZ4 and Snappy.

Check it out: github.com/minio/minlz

Speakers:Vincent Behar et Louis Fradin: Développeurs go chez UbisoftDescription:Talos est une distribution Linux qui se distingue par son caractère minimalis... Cloud Native Paris - Talos chez Ubisoft pour nos serveurs de jeux

Vincent Behar et Louis Fradin parlent de Talos et Omni à Cloud Native Paris. Écoutez l'enregistrement pour une démo et les détails techniques.

A Docker AI Agent story

When you (@rumpl.bsky.social) ask Docker's AI Agent to tell you a story about your running containers...

CloudNativePG @Mirakl & Talos @Ubisoft: Leurs retours d'XP!, jeu. 13 mars 2025, 19:00 | Meetup 🎉 **Bonjour à vous!** Pour notre prochaine rencontre venez assister à 2 supers retours d'expérience ✨! ❓ Si vous vous posez encore les questions du type: "Est-ce possibl

I'll be talking about Talos Linux and how we use it at Ubisoft at the next CNCF Paris Meetup: www.meetup.com/fr-FR/cloud-...