🌍 Dev.to

The arithmetic IS the product: building a planner that tells people the truth

Everyone planning their first pub quiz makes the same mistake. They count the questions, multiply by "about a minute", and announce a finish time. Then the night overruns by forty minutes, the kitchen has closed, and half the room left before the results. The error is not in the multiplication. It is that answering a question is the cheapest part of asking one. You read it out. Someone at the back says "what?". You read it again. A table argues. You wait. Then you read the answers out at th...
🌍 Dev.to

Google Introduces Gemini 3.8 Live Audio Models for Real-Time Voice AI Workflows

Google DeepMind has introduced Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking , two audio-focused models designed for more natural, near real-time conversations with AI. The models expand the Gemini Audio family beyond transcription, translation, and text-to-speech capabilities, with one aimed at high-volume voice interactions and the other built for more demanding reasoning while a conversation continues. For businesses, the development matters because voice can become a more pra...
🌍 Dev.to

Big Tech Issued $220 Billion of Bonds to Feed the Silicon Buildout

Through August 10 of this year, Alphabet, Amazon, Meta, Microsoft, and Oracle issued approximately $220 billion of corporate bonds, according to BNP Paribas data cited by Reuters. During the comparable period last year, that same group issued $12.5 billion. In less than twelve months, primary debt issuance from the five dominant cloud infrastructure operators expanded by a factor of seventeen. The immediate explanation is capital expenditure. S&P Global Ratings expects collective capex ac...
🌍 Dev.to

Your coding agent forgets on purpose. A PreCompact hook is where you save it

Your coding agent doesn't have amnesia. It has a budget. When the context window fills up, the host compacts it into a summary — and summaries keep the what while losing the why . For engineering decisions, the why is everything. This is a write-up of a tool I came across that is built around exactly that moment — the one where a decision can still be saved: myc , an open-source (MIT) local task-and-memory layer for coding agents, built on Bun and TypeScript. Its author describes three da...
🌍 Dev.to

Your Agent Eval Set Is Rotting: Build a Failure-Mining Loop for Google ADK

An agent evaluation set starts healthy. It contains the obvious intents, a few tool failures, and the happy paths used during development. Six months later, production has changed. New tools exist. Users phrase requests differently. A fallback introduced last quarter now handles 30% of traffic. Yet CI still runs the same twelve examples and reports green. The problem is not only stale prompts. The eval set itself is rotting. A durable evaluation program needs a controlled path from ob...
HTB - Archetype
🌍 Dev.to

HTB - Archetype

OS: Windows Difficulty: Very Easy This was a fun machine to do :) After the Nmap enumeration, we obtained the following results: Starting Nmap 7.99 ( https://nmap.org ) at 2026-09-15 11:05 -0400 Nmap scan report for 10.129.192.92 Host is up (0.24s latency). Not shown: 65523 closed tcp ports (reset) PORT STATE SERVICE VERSION 135/tcp open msrpc Microsoft Windows RPC 139/tcp open netbios-ssn Microsoft Windows netbios-ssn 445/tcp open microsoft-ds Windows ...
Quantity stepper UX: the number needs context
🌍 Dev.to

Quantity stepper UX: the number needs context

Tira’s shopping bag shows two products with a quantity of one. The free wipes get a small “Qty: 1” label. The face wash gets minus and plus buttons. Both numbers say the same thing. Only one is presented as something to edit. That detail is a useful starting point for quantity stepper UX. Before deciding how the buttons should look, decide what the number represents and what the customer can do with it. Three screens from Flownato’s reference library make that question concrete. ...
Why I Had to Learn About State Machines to Build a Desktop Pet
🌍 Dev.to

Why I Had to Learn About State Machines to Build a Desktop Pet

While building Mochi , I ran into a problem that a lot of us hit eventually: making something happen is easy. Controlling when it happens, how long it lasts, and what comes before and after is the hard part. It usually starts innocently enough: If an application opens, play this animation If the mouse moves here, make this happen And that works... for a while. Then the software grows up a little, and the questions get harder: What happens when a user clicks whi...
WFC in Perl: A Server-Driven UI Approach for Perl
🌍 Dev.to

WFC in Perl: A Server-Driven UI Approach for Perl

Perl has long been used for server-side web development, but modern web applications often rely heavily on JavaScript frameworks to manipulate the browser interface. WebForms Core (WFC) provides another approach: the server can generate UI commands that are executed by a browser runtime. This article demonstrates how WebForms Core can be used with Perl and the Mojolicious framework to build an interactive web interface while keeping the UI logic on the server. What is WebForms Core? ...
🌍 Dev.to

Solving IEEE-754 Floating-Point Errors in Client-Side Fintech: Introducing FinEngine

Modern fintech demands deterministic precision. In finance, there is no room for approximation—every minor fractional unit must remain 100% audit-compliant. Yet JavaScript, the backbone of client-side web applications, introduces structural challenges through IEEE-754 floating-point arithmetic. A single rounding drift can compound across amortization schedules and ledger balances. To address this challenge and establish local currency primitives, we launched FinEngine —an open-source comp...
🌍 Dev.to

Reef Connects Agent Feedback, Learning and Versioned Delivery

Human-Agent-Society has published Reef, open-source infrastructure that links agent inference, interaction feedback, learning and versioned delivery. For developers, the immediate consequence is that one system can manage updates to model weights or to an agent harness—including its prompts, rules and skills—rather than leaving learning and deployment as separate pipelines ( Reef repository ). One loop, two update surfaces Reef divides a learning cycle into four stages. Serve h...
🌍 Dev.to

UAJY Handbook RAG Chatbot Splits FAISS Search From Gemini

The public UAJY Academic Document RAG Chatbot repository packages a Streamlit assistant for the 2025/2026 academic handbook of Universitas Atma Jaya Yogyakarta’s Faculty of Industrial Technology. For developers, it provides an inspectable implementation of document ingestion, local vector search, cited answer generation, refusal controls and a 20-question evaluation harness in one codebase. From handbook PDF to cited answer The offline ingestion pipeline extracts text and table...
🌍 Lobsters

Performance Improvements in .NET 11

As is tradition with an imminent new release of dotnet, Stephen Toub has done a long writeup of what this means with regards to performance improvements. Comments
HackerNews

Can an AI chatbot replace human connection?

Article URL: https://www.scientificamerican.com/podcast/episode/are-ai-chatbots-making-us-less-human/ Comments URL: https://news.ycombinator.com/item?id=49715988 Points: 1 # Comments: 1
HackerNews

Meta launches subscriptions with enhanced AI features

Article URL: https://www.reuters.com/business/meta-launches-subscriptions-with-enhanced-ai-features-2026-09-15/ Comments URL: https://news.ycombinator.com/item?id=49716281 Points: 1 # Comments: 1
HackerNews

AI-Based Coding and Alienation

Article URL: https://arnerief.net/en/articles/ai-coding-alienation/ Comments URL: https://news.ycombinator.com/item?id=49716305 Points: 1 # Comments: 0
HackerNews

Measuring AI's Impact on Delivery Speed

Article URL: https://www.jamesshore.com/v2/blog/2026/measuring-ais-impact-on-delivery-speed Comments URL: https://news.ycombinator.com/item?id=49716333 Points: 2 # Comments: 0
HackerNews

Show HN: MyMarkdown – Make AI Markdown easier to read

I made a small free tool for humans (not AI) to read Markdown faster. If you build AI tools, you probably read a lot of Markdown. And you may have noticed that as it gets bigger, it gets harder for human eyes to digest. Here, just paste your Markdown, and it makes it cleaner and easier to read. It also makes JSON much nicer, which I needed a lot. It’s free. Maybe useful for other people building AI too. Comments URL: https://news.ycombinator.com/item?id=49716619 Points: 1 # Comments: 3