I have made thousands of static visualizations for Go, it's time to try something new: https://t.co/sQVmESgwXW

https://t.co/Wi4jbS2crv TL;DR Go Minimal language Very few features No syntactic sugar Lots of boilerplate Stable over years Easy to read Versatile slice type Automatic memory handling Team-oriented design Productivity focus Rust Feature-heavy language Zero-cost abstractions

VictoriaLogs v1.40.0 improves parsing speed for logs ingested via OpenTelemetry protocol by up to 10x (e.g. it reduces CPU usage by up to 10x)! Big thanks to @vadimaleksv for the pull request https://t.co/vRu7WIrsP3 !

Many articles make you think that go func() {} always gives you a goroutine with a 2 KiB stack. That's sometimes true, but not always. Go keeps a per-processor pool and a global pool of reusable goroutines, some with stacks already attached and some without. (diagram below)

Channels in Go are easy to get wrong unless you know the rules: - Sending to a channel panics if the channel is closed at any time before the send actually completes, even if the close happens "after" the send was started but before it has finished. - Receiving from a closed

Sometimes I receive private messages and emails from engineers who want to become a part of the VictoriaMetrics team. The best and the most reliable way to become our team member is to start helping other users with VictoriaMetrics products via GitHub issues, public Slack and

This repo lets you use the ? operator for error propagation in Go https://t.co/Lk96tI3icf A controversial project with the slogan: "Think TypeScript, but for Go."

Spotify successfully migrated from their internal observability system - Heoric - to VictoriaMetrics. This resulted in significant cost savings and improved user experience. https://t.co/07yHB1S7UC

Something big is gonna happen on this channel: https://t.co/eB7JlCvKr0. Wait for it!

How do std::Mutex and parking_lot::Mutex actually work under the hood? Which one is better?

How we avoided significant size growth for VictoriaTraces executable (from 20MB to 28MB, +40%) while adding support for receiving traces via gRPC. This also reduced CPU usage and improved performance for traces' ingestion over gRPC.

In #VictoriaTraces, we run an HTTP2 server to handle gRPC calls, instead of using the common gRPC-Go package for a gRPC server. Not trying to persuade others to do so, but we did find it fun. The general idea is to avoid including things you don't need. https://t.co/gO29ECkLQu

I really like this example because I prefer a simple writing style, extremely simple, even overly simple. But the wrong takeaway here might be using "Don't seek a library" for such a simple example

120+ diagrams done: memory, goroutine stack, garbage collection, all visualized. Coming soooon!

My book simplifies Go internals as much as possible, so it's very easy to understand. If you still don't get it, it's Go's fault 🙈

CPU Cache-Friendly Data Structures in Go: 10x Speed with Same Algorithm

Redrawing Chapter 7: How Go Memory Allocation Works. Still a long way to go, but really excited to be creative

I'm hiring for senior Go engineers to join us at @ona_hq! If you are: - AI native; using tools like Cursor, Claude code and Ona daily. - Are excited about building AI agents and the mission control platform that they will run on, entirely in Go. - Have 3-5 years of production Go

🔥 You can pass an expression to new() to directly get a pointer to its result, starting in Go 1.26. Before Go 1.26, you could only do new(Type) to get a pointer to the zero value of that type The new behavior new(expr) is syntactic sugar for: tmp := expr result := &tmp But

I wrote a post covering advanced Pattern Matching and best practices in Rust. It received a lot of likes on Reddit. Check it out here: