Lukaku Breaks Belgium's FIFA World Cup Scoring Record in Group G Win

Romelu Lukaku has become Belgium's all-time highest scorer at the FIFA World Cup after netting a record-breaking goal. The milestone was achieved during a dominant Belgium performance that secured a comfortable victory. The win proved crucial as it helped Belgium finish at the top of Group G in the tournament. Lukaku's historic goal marks a significant achievement in both his personal career and Belgian football history.

'India GDP growth may ease fiscal stress'

TokenBoard lets developers track AI coding token usage without sharing code or prompts

A developer has launched TokenBoard, a tool that tracks token usage and cost metrics for AI coding assistants without uploading any prompts or source code. The platform provides a local CLI that currently supports OpenAI Codex, reporting aggregate stats such as tokens used, sessions, messages, active days, and estimated cost. Users can view personal stats and optionally share a public stats card on a leaderboard. The creator plans to expand support to Claude Code and other AI coding agents in future updates. The project is in early stages and the developer is actively seeking feedback from Codex users on which metrics would be most valuable.

Dev Team Migrates Community Platform Luminary from Vanilla Stack to Next.js

Luminary, a community-powered directory and news platform spotlighting women driving change across various fields, has completed a two-week sprint migrating its codebase from vanilla HTML, CSS, and JavaScript to a modern framework. The project was restructured into a Turborepo monorepo housing a Next.js and TypeScript frontend alongside an Express and Supabase backend. The frontend now uses Tailwind for styling, TanStack Query for data fetching, and a shared component design system built on Base UI primitives. One contributor focused on converting the nominations and directory pages, refactoring DOM-driven forms into composable React components and splitting logic into dedicated service layers. The migration aimed to improve code consistency, scalability, and backend integration as the platform continues to grow.

Marathon through meadows: Baramulla’s rural run draws 1,000 participants

RHoiScribe Tool Aims to Stop AI from Breaking Hearts of Iron IV Mods

RHoiScribe is an open-source developer tool designed to help AI assistants work safely within Hearts of Iron IV modding projects. The core problem it addresses is that AI-generated mod code often looks syntactically correct but fails to load in-game, a gap between appearance and functionality common in HOI4 modding. Rather than feeding the AI a larger static knowledge base, RHoiScribe embeds it directly into the user's existing project environment, adapting to its structure instead of overwriting it. The tool includes project analysis features and a debugging workflow layer to make potential errors explicit before they cause failures. Future development will focus on deeper reference analysis, stricter validation rules, and more granular debugging workflows, with the project currently available on GitHub.

OpenAI and Broadcom Unveil Jalapeño, a Custom AI Inference Chip

OpenAI and Broadcom jointly announced Jalapeño on June 24, 2026, marking OpenAI's first custom Application-Specific Integrated Circuit (ASIC) designed exclusively for large language model inference. The chip pairs a reticle-sized compute chiplet with High-Bandwidth Memory, placing storage physically close to processing units to reduce the data-movement delays that typically bottleneck token generation. Unlike general-purpose GPUs, which handle training, graphics, and diverse workloads, Jalapeño is hardwired solely for inference, trading versatility for greater energy efficiency. Early testing indicates substantially improved performance-per-watt compared to GPUs, though final figures are still being measured. The design went from conception to tape-out in approximately nine months, which OpenAI describes as one of the fastest chip development cycles it has undertaken.

'Resign or join them': Raj Thackeray to top IPS officer over RSS praise

CodeBurn Tracks AI Spending Across 30 Tools With Per-Task Token Breakdown

CodeBurn is a new developer tool designed to provide detailed visibility into AI subscription and usage costs across platforms like Claude, Codex, and Cursor. The tool breaks down spending by individual task, model, and project, rather than showing a single aggregated bill. It supports monitoring of up to 30 AI tools from a single interface. All data is processed locally, meaning usage information does not leave the user's machine. Developers can run it instantly via the command 'npx codeburn' without a formal installation process.

Tamadoggo App Offers AI-Powered Pet Journal to Replace Cold Medical Trackers

Developer Sorin has launched Tamadoggo, an iOS app designed as a warm, story-like journal for pet owners rather than a clinical health tracker. Inspired by his own two dogs, Sorin built the app to log walks, meals, vet visits, weight, milestones, and everyday moments that accumulate into a readable life story. An AI layer runs quietly in the background, surfacing patterns, offering breed- and age-aware nudges, scanning vet documents, and generating monthly recaps without being intrusive. The free tier supports one pet with unlimited timeline entries and weight charts, while a paid Pro plan adds AI insights, document scanning, and support for multiple pets. The app is currently available on iPhone only, with Android support planned for a future update.

15-minute current booking rule boosts Vande Bharat train ridership across Southern Railway

Context Engineering: Why AI Systems Need More Than Just Good Prompts

As AI adoption grows in software development, engineers are moving beyond prompt engineering toward a broader discipline called context engineering. While prompt engineering focuses on what to ask an AI model, context engineering focuses on what information the model should have before it responds. Production AI systems like GitHub Copilot and Cursor combine multiple data sources — including system instructions, conversation history, tool outputs, and retrieved documents — to generate accurate responses. A well-contextualised request, rather than a cleverly worded prompt alone, is what drives reliable AI output in real-world applications. Understanding and applying context engineering is increasingly seen as a critical skill for teams building AI-powered products.

Voilaa Uses Google Gemini to Turn YouTube Videos into Interactive Learning Apps

A developer has built Voilaa, a full-stack educational tool that converts any YouTube video into a personalized interactive learning app using Google Gemini AI. Users paste a YouTube URL, select an academic depth level, and the system generates a self-contained HTML app featuring quizzes, flashcards, and data visualizations within seconds. The platform uses a two-stage AI pipeline: one model acts as a pedagogist to extract a curriculum blueprint and key terms, while a second model generates a complete single-file HTML application from that blueprint. Built with React 18, Node.js, and the Google Gemini API, the tool offers a three-tab workspace with a live preview, a Monaco code editor, and access to the underlying curriculum spec. Voilaa was submitted as part of the DEV Education Track challenge focused on building apps with Google AI Studio.

How a Developer's Mindset Can Transform Third-Party Risk Management

A software engineer turned risk professional argues that Third-Party Risk Management (TPRM) can be reimagined using familiar developer concepts like dependency audits, severity tiers, and structured data requests. Two common enterprise problems identified are shadow procurement—teams buying SaaS tools without oversight—and orphaned API integrations left active after projects end, both addressable through better visibility rather than stricter policy. The author recommends tiering vendors by risk level, similar to software severity ratings, and replacing vague compliance questions with specific, evidence-backed inquiries tied to frameworks like SOC 2, ISO 27001, or GDPR. Every identified risk gap should be tracked as a ticket with clear ownership, deadlines, and escalation paths, treating vendor-related breaches with the same rigor as internal incidents. Automation using lightweight custom scripts—rather than expensive enterprise suites—is suggested for tasks like vendor discovery, evidence collection, and risk scoring.

Inithouse Builds AI Pet Portrait Tool That Delivers 9 Art Styles in Under 60 Seconds

Inithouse, a multi-product software studio, has launched Pet Imagination, an AI tool that transforms user-uploaded pet photos into portraits across nine distinct art styles. The pipeline operates in four stages: image upload and preprocessing, prompt construction, AI generation, and portrait delivery. A fixed prompt skeleton with swappable style parameters ensures consistent output without maintaining separate prompt chains for each style. Generation times range from 18 to 35 seconds depending on style complexity, with the full process completing in under 60 seconds. No user accounts are required and uploaded photos are deleted from servers immediately after processing.

FacetLab: Free Open-Source WebGL Tool Lets Users Simulate Diamond Cutting

A developer has built and maintains FacetLab, a free browser-based WebGL simulator designed to teach the diamond 4Cs — cut, colour, clarity, and carat — through hands-on interaction. Users can inspect a rough crystal, select an orientation, polish facets, and receive a 4Cs scorecard summarising their results. The tool requires no signup or download and runs entirely in the browser. FacetLab is intended as an educational aid for jewellery beginners, STEM directories, and curious learners, and is not designed for professional appraisal or pricing purposes. The source code is publicly available on GitHub, and the live simulator can be accessed at 4cs.co.za.

Abhishek Sharma praises Sooryavanshi's curiosity as debut calls grow after T20I bench

Fifteen-year-old Vaibhav Sooryavanshi has drawn widespread attention following an impressive IPL 2026 campaign and a record-breaking List A innings. The young cricketer received his first T20I call-up for India's tour of Ireland but was left out of the playing eleven for the opening match. Teammate Abhishek Sharma highlighted Sooryavanshi's inquisitive attitude, noting that the teenager constantly asks questions to improve his game. India's batting difficulties in the first T20I have amplified public and expert demands for Sooryavanshi to be handed his international debut.

Google cuts Spanner cache costs 5% by treating memory as rent, not sunk cost

Google Research published a paper this week detailing a production deployment of linear elastic caching in Spanner, its globally distributed database. The approach, based on the classic ski rental problem, assigns each cached page a TTL at access time rather than relying solely on traditional eviction policies like LRU or LFU. In a fleet-wide rollout, the system reduced cache memory usage by 15.5% and lowered total cache ownership cost by around 5%, while cache misses rose by only 5.5%. To meet Spanner's performance demands, the team implemented the TTL predictor as a shallow decision tree translatable into a few lines of C++, using features such as page size, miss cost, and operation type. The key insight is that memory in a large fleet carries an ongoing cost, and pages should be evicted not just when space runs out but when the cost of retaining them exceeds the cost of a future cache miss.

India needs 2,000 GW new power capacity

GCL Season 4 Draft Set for Mumbai; Carlsen Returns, Divya Deshmukh to Debut

The Global Chess League (GCL) Season 4 Player Draft is scheduled to take place in Mumbai on June 29. The tournament itself will be held in Bengaluru from September 3 to 13. World No. 1 Magnus Carlsen is set to return to the league after sitting out last season. FIDE Women's World Cup champion Divya Deshmukh will make her GCL debut in the upcoming edition. The player pool comprises 36 Grandmasters, including chess legends Viswanathan Anand and Ian Nepomniachtchi.