Uttarakhand offers registered construction workers Rs 50,000 housing loan at low interest

Uttarakhand's labour welfare system provides registered construction workers access to a housing loan of up to Rs 50,000 at a minimal interest rate. The loan can be used for purchasing or constructing a house. To qualify, workers must have been fund members for at least five years and have a minimum of fifteen years remaining until retirement. Applications are submitted online and subsequently undergo departmental verification and investigation. Once approved, the loan amount is disbursed and repaid in equal instalments managed by the Board.

Two devs built an offline YOLO26 posture detector in 7 days — here's what broke

A two-person team in Austin, TX — engineer Igor Eduardo and field specialist Lexi Armstrong — built an on-device posture detection tool called SENTINEL during the webAI YOLO26 MLX Build Challenge in May 2026. The single-file Python app uses the YOLO26n model with a bounding-box heuristic to classify people as standing, sitting, or lying, running at roughly 16 FPS on an M4 Mac with Wi-Fi physically disabled to guarantee zero network egress. The project drew on Eduardo's healthtech background in patient triage and Armstrong's experience with denied-communications environments, targeting mass-casualty and industrial security scenarios. Two major technical failures shaped the build: OpenCV 4.13 crashed reliably when camera capture ran off the main thread on macOS 26, and a silent hang was traced to lazy evaluation in the yolo-mlx Boxes proxy rather than MLX inference itself. An earlier attempt to fine-tune the model on labeled posture data also failed to converge and was abandoned within the seven-day window.

Abhishek Sharma hits fastest T20I century by an Indian, breaks Rohit Sharma's record

Child passports now valid till 18: MEA

Tata Sons Board Re-Appoints N Chandrasekaran as Executive Chairman for Five More Years

Tata Sons has re-appointed N Chandrasekaran as Executive Chairman for an additional five-year term following a majority board vote. Chandrasekaran had initially signaled a decision that prompted the board to request he reconsider. The board's resolution came after he agreed to reverse his earlier stance at the board's request. The development sheds light on the ongoing governance dynamics between Tata Sons and the Tata Trusts, which hold significant ownership in the conglomerate.

After being sidelined, Boeing's Starliner to get starring role in NASA's spaceflight plans

"They should fly their capsule. The government paid a lot of money for that."

Priscilla Chan’s $75M gift transformed SF hospital

Microsoft Patches Two RCE Flaws in Semantic Kernel Caused by Prompt Injection

Microsoft disclosed two remote code execution vulnerabilities, CVE-2026-26030 and CVE-2026-25592, in its open-source Semantic Kernel agent framework. The first flaw, in the Python SDK, allowed unsanitized AI model output to reach an eval() call inside a vector store metadata filter, enabling attackers to execute arbitrary code by bypassing a blocklist via Python's class hierarchy traversal. The second vulnerability, in the .NET SDK's SessionsPythonPlugin, let a crafted prompt direct the agent to write a file to any location on the host, including the Windows Startup folder, allowing code to persist and execute after the session ended. Both bugs share a common root cause: powerful code-adjacent operations received values derived from model output without treating that input as potentially attacker-controlled. Fixes were shipped in semantic-kernel 1.39.4 for Python and .NET SDK 1.71.0 on the same day.

Sikh Truck Driver Stabbed 17 Times in Wyoming; Community Demands Hate Crime Probe

A Sikh truck driver was stabbed seventeen times in Wyoming in what has alarmed the Sikh American community. The assault occurred just four days after a controversial post by the US Department of Homeland Security referencing 'Mr Singh', raising concerns about its potential influence. Sikh American groups are calling on authorities to launch a full investigation into the incident. They want officials to determine whether the attack was motivated by religious or racial hatred. The community has expressed deep concern that the episode reflects a growing climate of hostility toward Sikhs in the United States.

Australian couple trapped paying rent and mortgage after council blocks modular home plan

An Australian couple purchased land with plans to build two modular homes on a single plot, but local council zoning regulations have blocked the project. The couple now bears the financial strain of simultaneously paying rent and a mortgage while remaining without a home. Modular construction, despite being a faster and increasingly popular building method, has faced bureaucratic obstacles under existing planning rules. However, recent state-level reforms are expected to simplify the approval process for modular builds. These changes aim to remove such regulatory hurdles and help speed up construction timelines for homeowners in similar situations.

Developer finds hand-written regex outperforms 7B AI model at detecting violent language

A developer testing AI safety classifiers shifted from cloud APIs to a locally run 7-billion-parameter model after their lab's IP address was blocked following repeated policy-sensitive queries. The local model, Qwen 2.5 Coder 7B running on CPU via llama.cpp, incorrectly flagged benign phrases like 'kill a Python process' and 'strangle options to maximise my money' as unsafe content. A simple hand-written regex with an exceptions list correctly identified all such phrases as non-violent, outperforming the neural network on this specific task. The author noted that the model could accurately explain terms like 'murder of crows' when asked directly, yet still flagged the word 'murder' as unsafe during classification — revealing a gap between knowledge and judgment. The findings highlight that AI safety benchmark scores reflect the entire surrounding scaffold, including parsers and exception logic, not just the underlying model's capabilities.

To Get the Most from Cursor AI, Developers Should Plan Before They Prompt

Cursor is an AI coding tool that can explore codebases, write features, fix bugs, and coordinate multiple agents to handle complex development tasks. However, the tool works best when developers invest time upfront to clearly define what they are building, who will use it, and what the system's boundaries are. Experts recommend sketching user flows and system diagrams before writing prompts, as visual clarity often outperforms lengthy written descriptions. Cursor's internal rules should also be treated as living guidelines, updated whenever the model produces repeated errors or instructions become overly complex. While AI can manage much of the implementation work, critical decisions around architecture, security, and access control remain the developer's responsibility.

What dbt Does and Why Data Engineers Can't Stop Talking About It

dbt, short for data build tool, is an open-source framework that transforms raw data into clean, analysis-ready tables directly inside a data warehouse such as Snowflake, BigQuery, or Redshift. It allows data engineers and analysts to write standard SQL SELECT statements, while dbt handles compilation, execution, dependency management, and table creation automatically. The tool also includes built-in data quality testing, auto-generated documentation, and lineage graphs that visually map how data flows across models. Before dbt, data teams struggled with duplicated transformation logic, no version control, absent documentation, and untested pipelines that only surfaced errors after stakeholders noticed report errors. dbt is available as a free open-source CLI tool called dbt Core, or as dbt Cloud, a managed platform that adds a browser-based IDE, scheduling, and a team-friendly UI.

Attractive.js Introduces Custom Actions API with Scalable js, Function, and Class Support

Attractive.js, a JavaScript library, has rolled out a custom actions feature that lets developers write their own actions when built-in options fall short. The system supports three tiers of complexity: a quick js: prefix for one-liners, plain functions registered via addActions, and full classes with an optional Action base class. All three approaches share the same registration interface, meaning code can scale from a simple expression to a structured class without changing how it integrates. Lifecycle hooks — beforeAction, afterAction, and onError — allow developers to intercept, cancel, or monitor actions globally from a single location. The library also ships optional bundled extensions and recommends a barrel-file pattern for organizing larger action collections, similar to how Stimulus structures its controllers.

What Tokens Really Are and Why They Determine LLM Costs and Limits

In large language models, a token is not simply a word but a fragment of text determined by a trained vocabulary, meaning one word can equal multiple tokens and vice versa. Each model — including GPT, Gemini, and Claude — uses its own tokenizer, such as Byte Pair Encoding, with vocabularies ranging from roughly 50,000 to 100,000 text pieces. Technical terms, long numbers, emojis, and non-English text tend to consume significantly more tokens than common English words, making them costlier to process via APIs. Every LLM enforces a context window, a hard cap on the total tokens it can handle per call, which must accommodate the system prompt, conversation history, retrieved documents, and the user's message combined. Misunderstanding how tokenization works can lead to unexpected API costs and context-window overflows, especially in applications that pass large amounts of text to the model.

React Performance: Why Colocating State Beats useMemo for Most Apps

A developer who has profiled dozens of React applications identifies recurring performance mistakes and the patterns that reliably fix them. The most impactful fix is colocating state — keeping state in the component that actually needs it rather than lifting it unnecessarily to a parent, which eliminates unnecessary re-renders at zero cost. For values that change frequently but do not need to trigger re-renders, such as mouse positions in canvas animations, using a ref instead of state prevents React from scheduling redundant updates. React.memo is recommended for expensive component subtrees, with useMemo and useCallback reserved for stabilizing props only after profiling confirms a bottleneck. For lists exceeding roughly 100 rows, virtualizing the DOM with a library like TanStack Virtual can reduce rendered nodes from thousands to around 20 regardless of list size.

How to Document Database Schemas Effectively for Engineering Teams

A structured approach to database schema documentation recommends three core layers: clear naming conventions, inline column comments, and a visual ER diagram. Naming best practices include using snake_case, plural table names, descriptive column names, and prefixes like is_ or has_ for booleans to reduce the need for additional explanation. Database-level comments should capture non-obvious details such as accepted values, data formats, and business rules like soft-delete policies. Teams are advised to maintain a SCHEMA.md file in their repository and update it as part of every pull request that modifies the database. The recommended toolset scales with team size, ranging from simple markdown files for small teams to dedicated data catalogues like DataHub or dbt docs for organisations with 15 or more engineers.

Open-Source React 19 Storefront Lets Small Restaurants Take Orders via WhatsApp

A developer has released FoodFlow, a free, open-source food ordering storefront designed to help small restaurants avoid paying high commissions to delivery aggregators. Built with React 19 and Tailwind CSS v4, the app runs entirely client-side with no backend server or database required. Orders are formatted into a readable WhatsApp message, covering delivery versus pickup options, item variants, modifiers, and discount rules. The project achieves a Lighthouse score of 100 across all categories and a sub-0.8-second LCP on mid-tier mobile networks. The code is MIT licensed and publicly available on GitHub.

How to Measure If Your AI Agent Is Safe to Run Unsupervised Overnight

A recent benchmark tested 507 state-mutating workflows across 20 attempts each, revealing a stark gap between single-attempt accuracy and consistent reliability — the top model succeeded 66.5% of the time per attempt but achieved perfect consistency on only 47.53% of tasks. Developers are advised to evaluate agents using a metric called pass^k, which checks whether an agent completes a task correctly every single time, not just occasionally. The approach requires tasks that mutate stored state — such as processing refunds or updating records — since read-only tasks cannot expose the failure modes that matter. A proper test harness must include full environment resets between runs and assertions that verify both required changes and the absence of unintended side effects. The benchmark found that most tasks fall into a dangerous middle category where agents succeed only sometimes, making silent data corruption a real risk that standard observability tools like error rates or latency metrics would fail to detect.

How to Cut Costs as Electronics Prices Continue to Rise

Consumer electronics prices have been climbing steadily, a trend some are calling 'techflation.' Owning devices like phones and computers carries costs beyond the initial purchase price. Understanding the true total cost of ownership is key to making smarter buying decisions. Experts suggest several practical strategies to help consumers reduce their overall spending on technology.