SShortSingh.

Programming

0
ProgrammingDEV Community ·

Four Warning Signs It's Time to Drop a Client — and Revenue Isn't One

A services business analysis argues that using revenue as the sole metric for dropping clients is counterproductive, often leading firms to retain high-paying but disruptive accounts while losing easier, more collaborative ones. The real costs of a difficult client go beyond logged hours, manifesting as fragmented team attention, senior staff avoiding certain projects, and slower output across all other accounts. These hidden costs rarely appear on a profit-and-loss statement, instead quietly degrading team performance and account health over several months. The piece outlines four early warning signals that indicate an account has become unsustainable, appearing in a specific sequence before financial damage becomes visible. The goal is to give service businesses an operational framework for identifying and acting on problem accounts before the damage compounds.

0
ProgrammingDEV Community ·

IBM Bob AI Assistant Tested on Real-World IBM Maximo Tasks by Developer

Developer Yann Bordelanne evaluated IBM Bob, an AI assistant, against practical IBM Maximo challenges faced by developers, administrators, and consultants. His findings showed that richer context — including business rules, naming conventions, and existing scripts — significantly improved the quality and relevance of Bob's outputs across SQL generation, XML configuration, BIRT report editing, and automation scripting. Bob also demonstrated utility in accelerating early-stage troubleshooting by helping identify Java library conflicts and class-loading issues from error logs. Looking ahead, Bordelanne highlighted the Model Context Protocol (MCP) support in Maximo Application Suite 9.2, which could allow Bob to interact directly with live Maximo data — such as querying work orders or recording meter readings — while remaining governed by existing security controls and business logic.

0
ProgrammingDEV Community ·

Developer Unifies SEO, SMM, and Traffic Data Into a Single CRM Dashboard

A small business owner replaced three separate marketing reports with a unified dashboard built over 14 years of fragmented promotion workflows. A custom collector script pulls traffic analytics, a 1,188-phrase semantic core, sitemap coverage data, and SMM publication statuses into a single JSON feed refreshed in under a minute. Semantic coverage currently stands at 80%, with gaps automatically ranked into a content plan, removing guesswork from editorial decisions. An AI-assisted content pipeline adapts each article for five platforms — including Telegram, VK, Dev.to, and LinkedIn — and runs a five-point linter to filter machine-like writing before human review. The developer plans to add a Google Search Console adapter next, closing the loop between published content and actual search query performance.

0
ProgrammingDEV Community ·

Four Common Pitfalls That Cause PDF Lease Forms to Silently Drop Filled Values

When filling PDF lease forms programmatically, values can appear correct in code yet fail to display in the final document due to field name changes, hierarchy mismatches, or missing appearance streams. PDF forms are structured as trees with fully qualified field names, meaning a template revision can rename a field while leaving its visible label unchanged, causing code to silently skip the write. Bundle operations such as splitting and merging multi-page lease packets compound the problem, as shared field names across source documents can cause widgets to overwrite each other or lose data. Developers are advised to enumerate terminal field names and types, compare them against their value mappings, and save-then-reopen the output before rendering to catch serialization losses. A two-stage validation approach — structural checks on every revised template followed by targeted page rendering — is recommended to balance data fidelity against the high cost of full-document rasterization.

0
ProgrammingDEV Community ·

How to Build a Safe, Lightweight Image Pipeline for HTML Catalog Emails

A recommended pipeline for healthtech catalog emails places background removal, content moderation, format conversion, and compression in strict sequence before any image reaches an email template. Moderation must function as a mandatory release gate rather than an optional step, with approval status recorded against each asset identifier. Images should be converted to widely supported formats, compressed aggressively, stored privately, and referenced in HTML via time-limited presigned URLs instead of being attached directly. Email clients are among the least capable image renderers, making small file sizes in familiar formats a higher priority than modern or visually rich formats. When evaluating image API providers such as Cloudinary, ImageKit, or Sharp, teams are advised to assess moderation coverage first and transformation convenience second.

0
ProgrammingDEV Community ·

Why Semantic Layers Matter for Making Complex SQL Usable by Humans and AI

Large analytical SQL queries often span hundreds of lines, distributing meaning across nested CTEs and complex joins in ways that make them difficult for analysts and AI systems to interpret reliably. The core challenge, known as semantic compression, involves translating physical data complexity into a structured representation that conveys business meaning rather than just computation. Snowflake defines semantic views as schema-level objects that model business entities, metrics, and relationships on top of raw data, effectively separating what a value means from how it is calculated. Research benchmarks like Spider and RAT-SQL have shown that text-to-SQL models perform significantly better when given strong semantic context, rather than being forced to infer meaning from raw database schemas. Experts caution that simply wrapping an entire complex query inside a semantic view is insufficient; the semantic layer should expose business-defined meaning while leaving implementation details in the physical layer.

0
ProgrammingDEV Community ·

Developer Builds AI Support Agent With Persistent Memory Across Conversations

A developer has built a stateful customer support agent that retains memory of past interactions, solving a key limitation of traditional chatbots. Most conventional support bots are stateless, requiring customers to repeat information such as error codes and ticket details each time they return. The new system uses a memory layer called Hindsight, which stores and retrieves customer history through semantic, keyword, graph, and temporal recall strategies. Before generating any response, the agent first pulls relevant historical context — including prior tickets, error messages, and troubleshooting steps — and feeds it to the language model alongside the current query. The architecture follows a three-step memory loop of recall, reasoning, and retention to maintain continuity across multiple support conversations.

0
ProgrammingHacker News ·

Vespper Launches MCP Tool Letting AI Agents Edit Word Docs via HTML

Y Combinator-backed startup Vespper has launched an MCP (Model Context Protocol) server that enables AI agents to edit Word documents more efficiently by presenting them as HTML rather than raw OOXML. Founded by Dudu and Topaz, the team identified the problem while building an AI document editor for pharmaceutical companies, where filling a single clinical study report took around 50 minutes with poor results. Existing approaches — such as using low-level libraries, opinionated MCP tools, or Markdown round-tripping — either waste context tokens on Word mechanics or lose document fidelity. Vespper's solution converts DOCX to HTML as a read-only projection for the agent, then reconciles edits back into the original file, keeping the source document intact. The company claims its tool is currently three times faster, twice as cheap, and more accurate than the nearest alternative.

0
ProgrammingDEV Community ·

Vireo AI Launches Smart Productivity Platform Powered by OpenRouter and Cloud VPS

Vireo AI is a new AI-powered productivity platform built on the OpenClaw autonomous agent framework, designed to optimize daily workflows through intelligent task management. The platform uses OpenRouter as a unified API gateway to access large language models, particularly Google Gemini, allowing flexible model selection without altering backend code. A key feature called Live Sync AI Chat Bubble enables real-time, seamless communication between users and autonomous AI agents without page reloads. The system processes free-text instructions through deep semantic extraction rather than simple keyword matching, automatically storing structured outputs in a persistent SQLite database. Vireo AI is deployed on Cloud VPS infrastructure and its source code is publicly available on GitHub.

0
ProgrammingGitHub Blog ·

Git 2.56 Released: Key Features and Changes Highlighted by GitHub

The open source Git project has officially released version 2.56 of its widely used version control system. GitHub published a roundup highlighting the most notable features and changes introduced in this latest release. The update continues Git's tradition of iterative improvements contributed by its open source community. GitHub's overview serves as a guide for developers looking to understand what has changed since the previous version.

0
ProgrammingHacker News ·

Pew Research Center outlines its guidelines for AI use in research work

Pew Research Center has published a transparency statement explaining how it does and does not use artificial intelligence in its work. The organization outlined specific boundaries and approved applications for AI tools within its research processes. The disclosure reflects growing pressure on research institutions to be open about AI's role in producing findings. By clarifying these boundaries, Pew aims to maintain public trust in the credibility and methodology of its studies.

0
ProgrammingDEV Community ·

Developer Builds Incident-Recovery Agent That Validates Memory Before Applying Fixes

A developer built AfterTrace, a command-line incident-recovery tool, after their AI agent nearly applied a wrong fix by relying on a past solution for a similar-looking but different bug. The agent had previously fixed a failing search query by correcting a misconfigured alias, but in a repeat incident the alias was already correct and the real culprit was a stale cache. To prevent such errors, AfterTrace follows a principle where recalled fixes only guide what to investigate first, not what to change. Before applying any stored fix, the agent re-checks the current live state to confirm the fix still applies, rejecting it outright if conditions have changed. The developer plans to expand fault-type coverage and log rejected recalls so the agent can also learn from its own corrections.

0
ProgrammingDEV Community ·

Google's goto Redirect Wraps Organic Links in US, Affecting Analytics Not Rankings

Google has nearly completed rolling out google.com/goto redirect wrappers for organic search result links in the United States as of late August 2026, adding an intermediary Google URL before users reach their destination. The change does not alter how pages are ranked; it only modifies the click path a user travels before landing on a website. However, the extra redirect hop can cause discrepancies in analytics tools, rank trackers, and SERP APIs that do not correctly follow the redirect to its final destination URL. Google Search Console is unaffected in terms of its core reporting logic, as it already groups performance data by the final resolved URL after redirects. Site owners are advised to audit their measurement workflows — comparing GA4, Search Console, and third-party SEO platforms — to ensure each system is consistently recording the final destination rather than the intermediary Google URL.

0
ProgrammingDEV Community ·

Apache Foundation Completes AI-Powered Security Scans Across 230 Repositories

The Apache Software Foundation (ASF) conducted full security scans across 230 of its code repositories during a three-day window in August 2026. The work was carried out jointly by the ASF Security and ASF Tooling teams under the Foundation's Responsible AI Initiative, using Anthropic's Claude Mythos 5 via a program called Project Glasswing. The effort was partly motivated by a rise in low-quality AI-generated vulnerability reports reaching open-source projects, prompting ASF to run its own rigorous, context-aware scans internally. Scanning was powered by an automated pipeline built on the Gofannon agent platform, evaluating code against the OWASP Application Security Verification Standard across three model tiers of increasing analytical depth. Findings are now being disclosed to the respective project teams through official ASF channels, with remediation already underway, and the initiative was made possible in part by a donation from Anthropic.

0
ProgrammingDEV Community ·

Why Encryption Needs a Unique Nonce Every Time, Even With the Same Password

In modern symmetric encryption, a nonce — a randomly generated value mixed into each encryption call — ensures that identical messages encrypted with the same key still produce different ciphertexts every time. Without this, an attacker monitoring encrypted traffic could detect patterns and infer information without ever cracking the key itself. In AES-GCM, a 12-byte nonce generated via os.urandom() is prepended to each encrypted output and passed back during decryption, where its role is uniqueness rather than secrecy. Reusing a nonce even once under the same key can fully compromise the confidentiality of affected messages and potentially expose the authentication key used for tamper detection. Nonce reuse is among the most common real-world implementation flaws in otherwise well-designed encryption systems, making proper random generation on every encrypt call a non-negotiable requirement.

0
ProgrammingDEV Community ·

Why Passive Tutorial Watching Stalls Game Dev Progress in GDevelop

Many aspiring game developers using GDevelop fall into 'tutorial hell,' spending hours following courses without producing anything meaningful of their own. The core issue is passive, 'just-in-case' learning, where skills are absorbed without an immediate project to apply them to, causing knowledge to fade quickly. Experts recommend switching to 'just-in-time' learning — starting with a personal game idea and searching for solutions only as specific problems arise. Learning from creators who are only slightly more advanced can also be more effective, as they use current tools and avoid skipping steps that beginners need. Finally, teaching or explaining what you have built, even to yourself, helps identify knowledge gaps and reinforces understanding faster than any structured course.

0
ProgrammingHacker News ·

Driver Wrongly Ticketed After Flock's AI System Falsely Flagged Her as Uninsured

A driver was issued a ticket for having no insurance despite actually being insured, after automated license plate recognition company Flock Safety incorrectly flagged her vehicle. Flock Safety, a Y Combinator-backed startup founded in 2017, provides surveillance and data services to law enforcement agencies across the United States. The incident highlights growing concerns about the reliability of AI-driven systems when used as the basis for police enforcement actions. The driver faced legal consequences stemming solely from an algorithmic error, with no independent verification by officers before the ticket was issued. The case raises broader questions about accountability and due process when automated systems produce false results that directly affect citizens.

0
ProgrammingDEV Community ·

How Polymarket Order Book Data and Chainlink TWAP Can Estimate Fair BTC Probabilities

A quantitative research approach proposes combining Polymarket order-book prices with Chainlink TWAP data to better estimate fair probabilities in short-term Bitcoin prediction markets. Polymarket's BTC Up/Down markets resolve based on Chainlink's BTC/USD Data Stream rather than arbitrary spot exchange prices, making the two price signals distinct and non-interchangeable. The framework separates market-implied probability derived from order-book data from the underlying reference price supplied by Chainlink, treating them as complementary layers of information. A simplified Black-Scholes binary option model is suggested as a research tool to generate a probability distribution, which can then be compared against the market price to identify a potential model edge. Researchers caution that the Black-Scholes component depends heavily on volatility assumptions and should not be treated as a definitive oracle for market resolution outcomes.

0
ProgrammingDEV Community ·

Developer Builds AI Whiteboard That Converts Hand-Drawn Sketches Into Live Diagrams

Fullstack developer Taqui has launched Drawva, an infinite-canvas whiteboard application powered by a multimodal AI agent built on a DeepSeek-based microkernel runtime. The tool interprets freehand strokes in real time and converts them into interactive diagrams, code widgets, math equations, and live data visualisations. Drawva supports 11 runtime plugins — including weather, stocks, and GitHub metrics — and seven sandboxed diagram formats such as Mermaid, Graphviz, and GeoJSON. Custom prompt-engineering contracts ensure the AI preserves the user's original pen strokes and avoids duplicating handwritten content. The app saves data locally via IndexedDB and enables peer-to-peer collaboration through WebRTC, with no central server database required.

0
ProgrammingDEV Community ·

Developer Reflects on Nine-Protocol Journey Into Industrial Cybersecurity Research

Security researcher RUGERO Tesla, writing on DEV Community, concluded a nine-part series on researching industrial protocols used in ICS/OT environments. Tesla described how early attempts involved oversized lab setups and unfocused exploration, but the approach gradually shifted toward smaller, targeted experiments and packet-level evidence gathering. The researcher noted that an interest in building and breaking controlled systems, rather than hunting bugs, drew them toward industrial security. A turning point came when building a local Modbus test environment to safely develop an exposure-analysis tool, replacing an initial plan to probe live systems found on Shodan. That decision to work within self-controlled environments became the foundation of a repeatable research methodology applied across all nine protocols in the series.

← NewerPage 836 of 3172Older →