VarBear

Pinterest stitched a sane dev console while Duolingo weaponized FinOps; meanwhile, the rest of us are juggling Linux power moves, modern SQL, an actually-useful research agent, and a kernel built from scratch. Java refuses to fade, AI is turning seniors into ‘babysitters,’ and a detour into esoteric languages keeps the curiosity sharp—dive into the details below.

🐧 Best 20 Linux Commands for Daily Use in Production Servers
🧩 Developer Experience at Pinterest: The Journey to PinConsole
🌀 Esoteric Languages Challenge Coders to Think Way Outside the Box
💸 How FinOps Drives Value for Every Engineering Dollar
🤖 How to Build an Advanced AI Agent with Search (LangGraph, Python, Bright Data & More)
☕ Is Java Still Used? Current Trends and Market Demand in 2025
🌳 SQL needed structure
🔄 The Ultimate Sprint Retro: My 10 Years of Software Engineering
🍼 Vibe coding has turned senior devs into ‘AI babysitters,’ but they say it’s worth it
🧪 Writing an operating system kernel from scratch


Less noise, more leverage—go build.

Have a great week!
FAUN.dev Team

Code. Game. Flow. This 9"×8" (22.86 x 20.32 cm) Binary Matrix mouse pad gives you smooth precision, durable build, and a design every developer will vibe with. Perfect for work or play.

👉 Get yours today for €12,95 – ships in 2-9 days.

Pinterest rolled out PinConsole, a custom-built Internal Developer Platform powered by Backstage. Years of scattered tools had piled on complexity. This is their clean slate.

PinConsole pulls developer workflows into one place, plugging into PinCompute (Kubernetes), GitHub, Jira, and PagerDuty. It also brings a unified data model to the table. The payoff? Faster onboarding. Less mental tax. A developer NPS north of 70.

Daniel Temkin has written a book about 44 esoteric programming languages, including Valence, which uses ancient Greek measuring symbols. Temkin emphasizes the significance of esoteric languages in promoting creativity and investigating the complicated nature of modern programming. These languages have a long history dating back to the early days of computing, with examples including INTERCAL (1972) and Befunge (1993).

Fastly says 95% of developers spend extra time fixing AI-written code. Senior engineers take the brunt. That overhead has even spawned a new gig: “vibe code cleanup specialist.” (Yes, seriously.)

As teams lean harder on AI tools, reliability and security start to slide—unless someone steps in. The result? A quiet overhaul of the dev pipeline. QA gets heavier. The line between automation and ownership? Blurry at best.

Java’s not just hanging on in 2025—it’s running the show. Over 90% of the Fortune 500 still trust it to power cloud platforms, big data pipelines, and IoT sprawl.

What’s keeping it sharp? A brisk six-month release cadence. A battle-hardened ecosystem through OpenJDK and Jakarta EE. And a JVM that keeps leveling up—think smarter JIT compilation and tighter garbage collection.

The big picture: Java’s fully plugged into the new stack—microservices, cloud-native gear, and scale-hungry workloads. The old lion didn’t retire. It evolved.

A decade in the trenches took one engineer from writing clean code to navigating company chaos—eventually landing in engineering management. The big shift? Less about scaling systems, more about scaling humans.

What started with system design and production code morphed into leading teams, syncing with product, and driving business results. The turning point hit during a rough patch in tech—layoffs, uncertainty, and tighter margins. That’s when tactical, technically fluent managers stopped being “nice to have” and started running the show.

The big picture: Engineering leadership isn’t just about code anymore. It’s a hybrid job—part debugger, part compass.

Raku throws together multi-paradigm support, gradual typing, first-class regex grammars, and metaprogramming that actually earns the name. It comes with built-in concurrency, multiple dispatch, and fresh tools like RakuAST for syntax-aware code wrangling.

Duolingo’s FinOps crew didn’t just track cloud costs—they wired up sharp, automated observability across 100+ microservices. Real-time alerts now catch AI and infra spend spikes before they torch the budget.

They sliced TTS costs by 40% with in-memory caching. Dumped pricey CloudWatch metrics for Prometheus. Same insights, lower bill.

A fresh roundup drops 20 go-to Linux commands for production sysadmins, dialing in on modern defaults like htop > top, ss > netstat, and ip > ifconfig. The shift? Faster tools that actually get updates. Built with systemd in mind, too.

Expect the usual suspects—journalctl, rsync, crontab—all still pulling weight for logs, file sync, and scheduled jobs.

Modern SQL can now shape hierarchical data straight from the query. One query, one response—no post-processing, no glue code.

Forget juggling ORMs to patch the mismatch between flat tables and nested UI data. No more extra trips to the database. No risk of half-built transaction states.

A barebones time-sharing OS kernel, written in Zig, running on RISC-V. It leans on OpenSBI for console I/O and timer interrupts. Threads? Statically allocated, each running in user mode (U-mode). The kernel stays in supervisor mode (S-mode), where it catches system calls and context switches via timer ticks.

One neat trick: kernel and userland share a single binary. No dynamic linking. No loaders. Everything stitched together upfront.

In this video, you'll be building an advanced AI agent in Python using LangGraph - a multi-step, deep research agent that will pull live data from sources like Google, Bing, and Reddit. The video is covering advanced Python concepts, complex architecture, and best practices for building agents that go far beyond just a single prompt response.

Scenario-driven simulator for async distributed systems: declare topologies, run simulations, measure latency/throughput, resource usage and see where bottlenecks emerge.

Modern Backend Framework that unifies APIs, background jobs, workflows, and AI Agents into a single core primitive with built-in observability and state management.

 Opinionated Arch/Hyprland Setup

Codefather protects your codebase by controlling who can change what. Set authorization levels, lock down files, and enforce your rules—offline via CLI or online with Actions.

Did you know that the Garbage First (G1) collector in OpenJDK marks any object larger than half a memory region as humongous and places it directly in the old generation, skipping the young heap? These objects are never compacted during normal collections and are only released when entire regions can be freed, which can quietly waste memory and increase pause times. With 1 MB regions, even a ~600 KB byte[] counts as humongous, so a common fix is to raise the region size or split buffers into smaller chunks.

"Feature flags reduce risk by multiplying states; you're safer and more confused. If you don't delete them, they become your architecture."
— SenseiOne

This Week’s Human is Gursimar Singh, a Google Developers Educator, Author @ freeCodeCamp, and DevOps & Cloud consultant who makes complex systems teachable. They’ve spoken at HAProxyConf 2022, multiple KCDs, and DevOpsDays Warsaw; reviewed programs for OpenTofu Day and PyCon India; and mentored at IIT Madras while volunteering with EuroPython. Offstage, they’ve published 70+ articles reaching 100k+ readers and contributed 5 project write-ups to the Google Dev Library, covering tools from Kubernetes to Terraform.