Developer News

Comments

Learning how to add evals to an LLM feature is the difference between shipping a demo and shipping a reliable product. When you embed an LLM into a real feature - a chatbot, a voice agent, a document summarizer - you’re not just calling a model. You’

Every Java engineer who has connected JConsole - or JDK Mission Control - to a server in another network segment knows the ritual. Open the JMX port. Discover that RMI quietly opened a second port - random by default. Pin it with a system property no

Welcome to Learning Xahau, a series of articles dedicated to helping developers, builders, and blockchain enthusiasts better understand the Xahau ecosystem. Whether you're just getting started or already building advanced applications, these posts wi

Developers flagged a freshly filed, reproducible issue that should make anyone running background agents pause: Claude Code's background Opus subagents intermittently stall on their first turn and, instead of producing useful work, emit system-prompt

If you've ever tried to duplicate a Keycloak realm on the same server - say, to spin up a myrealm-dev realm alongside your existing myrealm - you've probably hit this wall: Export the realm from the Admin Console ( Realm settings → Action → Partial e

Lukeed's sirv is legendary, but the edge runtime landscape has shifted. We need zero-copy streaming, native Web APIs, and composable transforms without pulling in Node-specific shims. Enter @rabbx/sirv . It’s an ultra-fast, zero-dependency static fil

OpenClaw took a clear step this window from "agent you run locally" toward "agent you run as infrastructure." Two commits do the heavy lifting. A Claude session fleet The first adds a Claude session fleet . Rather than one Claude session per user or

Most engineering teams I talk to can ship an AI demo. The prototype works, stakeholders are impressed, and everyone agrees The post The impressive AI demo is dead. Here’s what actually reaches production appeared first on The New Stack .

Yesterday's release post was about the bigger business line we will not cross: no royalties on IAP, ads, commerce or app revenue. This post is about one of the most concrete platform additions in that release. PR #5281 adds Apple CarPlay and Google A

This morning our corner of the agent economy grew a couch. We run a working system of 15 AI agents: a legal bot with live clients, an accountant, a content engine, an ops dispatcher. Operate that many agents and you learn something uncomfortable: mos

Inside the Black Box - Post 2 of 6 Series: Inside the Black Box - A developer's honest guide to how AI actually works, what's broken, and where it's all going. Post 1: Nobody Knows Why It Said That Let me paint a picture. You're hiring for a senior e

Hey everyone! If you've ever tried building a desktop application in Go, you probably know the drill. You either end up using Electron (and suddenly your simple utility app eats 500MB of RAM), or you try a native GUI framework only to get stuck in a

You tweak the prompt. Run it against the three examples you always use to sanity-check. It looks better. Ship it. That's not evaluation. That's vibes with extra steps - and it's the exact habit interviewers are trained to catch with one question. 🎯 T

Launching a Product: A Translation Problem Launching a product is a translation problem. Hacker News punishes anything that smells like marketing. X gives you 280 characters. The Play Store gives you 30 for a title and 80 for the short description, a

Proton's Lumo 2.0 finally looks like a real ChatGPT rival with reasoning, image generation, web search, and memory baked in.

Go has emerged as the lingua franca for cloud infrastructure, used for everything from container orchestration and CI/CD pipelines to The post Microsoft joins Google in backing Go for AI agents - OpenAI and Anthropic lag appeared first on The New Sta

We've all been there, especially as developers or first-time founders. We keep adding one more feature, one more screen, one more setting, or one more animation because we believe more equals better. It feels productive. It feels like progress. But m

TL;DR- Inference cost conversations obsess over FLOPs and token prices, but the real constraint on LLM serving is memory bandwidth- specifically the cost of moving the KV cache in and out of GPU memory on every decode step. Teams that optimize for co

A deep dive into the first stage of every card payment. Most of us have made a card payment and seen the familiar message: Transaction Successful The entire process takes only a few seconds. From the customer's perspective, that's the end of the stor

You are not really deciding whether a tracker is good today. You are betting on whether it will still be the tracker you want a year from now, after the funding round, the acquisition, the pivot to an AI platform, or the quiet shutdown email. That be

I got tired of a routine most of us know too well: you need to merge two PDFs, or shrink an image, or make a QR code - so you Google it, land on some ad-covered site, and it asks you to upload your file to their server and maybe "create a free accoun

The Setup Blueprint Constructing a reliable Python pipeline in VS Code requires anchoring the system with isolated environments. After deploying the editor and core extensions, the workflow shifts to generating a distinct environment layer. Linking t

Over the past few months, I've experimented with several workflows for using LLMs in software development. In this article, I describe the local setup I've validated on my own PC: Ollama running on Windows, Aider inside Ubuntu on WSL2, and a code-foc

Every AI coding tool now has a "project rules" file. Cursor has .cursor/rules , Claude Code has CLAUDE.md , OpenAI Codex has AGENTS.md . Teams write them once, watch the agent ignore half of it, and conclude the feature is broken. Most of the time th