SShortSingh.
TopTechnologyAI & MLStartupsProgrammingScienceBusinessFinanceCrypto & Web3World

Programming

LatestTopDiscussed
0
ProgrammingDEV Community ·

Developer details browser-based tool for meeting strict government photo upload limits

Government visa forms like the US DS-160 and China's online system enforce strict JPEG file size windows, often between 54KB and 240KB with compression ratio limits. A developer has created a browser-based tool that automatically adjusts photo quality to fit these constraints without server involvement. The tool uses a binary search algorithm on JPEG quality settings to find the largest acceptable file size within the required range. It prioritizes higher quality within the limit to reduce rejections for blurry photos and handles edge cases where photos are inherently too large or too small.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

TypeSafe AI launches JEV model for faster, cheaper AI agent decisions.

TypeSafe AI has introduced a new AI model named JEV, designed specifically for software decision-making. Unlike generative large language models that produce text, JEV evaluates predefined options to output structured, software-actionable decisions. The company reports that JEV can be significantly faster and cheaper for select workflows compared to using LLMs for similar tasks. This approach is positioned as optimal for the frequent, constrained choices required by AI agents, such as routing tasks or selecting tools.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Flawed test data led to duplicate charging station listings on map

In July, a developer discovered duplicate listings for the same physical charging sites on their company's map. The map aggregates data from sources like the German federal register and roaming networks, using a nightly job to deduplicate records. A May refactor of this job introduced a test suite that incorrectly assumed operator names from different sources would match. This flawed test data, not based on real-world examples, allowed most real duplicates to remain, as operator names rarely align between sources.

0 comments Read more at DEV Community
0
ProgrammingHacker News ·

Complex brain wave patterns provide new insights into neural activity

Scientists have discovered surprisingly intricate wave patterns within the brain. These patterns are offering researchers a clearer view into the organ's internal mechanisms and functions. The findings, reported by Quanta Magazine, stem from recent neurological research. Analyzing these waves helps explain how large networks of neurons coordinate activity. This work advances the fundamental understanding of brain operation.

0 comments Read more at Hacker News
0
ProgrammingDEV Community ·

How Mojo Uses SIMD to Speed Up Text Search Across Large Documents

SIMD, or Single Instruction Multiple Data, allows a processor to compare multiple bytes simultaneously rather than one at a time, making it far more efficient for large-scale text searches. The Mojo programming language exposes this CPU-level capability through a typed SIMD construct where both the data type and number of lanes are part of the value's type signature. A tutorial on DEV Community walks through building a Mojo application called ResearchLens, which uses SIMD to quickly identify candidate word matches in research documents. The approach pairs a fast SIMD scan for first-byte matches with a scalar verifier to confirm full word matches and handle edge cases like partial vectors. No GPU is required, as the SIMD hardware used is already present in modern CPUs.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Why HPC Clusters Use InfiniBand: Low Latency and RDMA Explained

High-performance computing (HPC) clusters often deploy InfiniBand alongside standard Ethernet because tightly coupled workloads demand more than just high bandwidth. Applications running across dozens or hundreds of nodes continuously exchange data, making both latency and throughput critical to overall performance. InfiniBand's key advantage lies in Remote Direct Memory Access (RDMA), which enables data transfers directly between nodes' memory with minimal CPU and operating system involvement, reducing communication overhead. However, InfiniBand is not the only option — high-speed Ethernet and RoCE, which brings RDMA capabilities to Ethernet, are also used depending on workload requirements. Ultimately, the network is as vital to HPC performance as the CPUs themselves, since powerful processors are ineffective if nodes spend excessive time waiting for data.

0 comments Read more at DEV Community
0
ProgrammingHacker News ·

Modular Industrial Dashboards Predated Digital Pixel Displays

An article on Hacker News examines the era before digital screens, focusing on modular industrial dashboard systems. These physical control panels, built from individual gauges and indicators, were used to monitor and manage machinery. The piece explores their design, functionality, and historical significance in industrial settings. The discussion highlights how these analog systems laid groundwork for modern human-machine interfaces. The original article and community comments were hosted on the Hacker News platform.

0 comments Read more at Hacker News
0
ProgrammingDEV Community ·

WebDecoy Plugin Lets Fastify APIs Detect Scrapers and Bots in Monitor Mode

Developers can add request detection to Fastify APIs using the open-source WebDecoy plugin, which identifies whether traffic comes from frontends, integrations, or scrapers hitting the same endpoint. The plugin operates in a non-blocking monitor mode, meaning suspicious requests are flagged and logged but not blocked, leaving API responses unchanged. A tripwire rule can be configured on sensitive paths such as /.env to trigger a DENY verdict whenever those routes are probed. Detection results are logged locally by default, and connecting an API key unlocks a cloud dashboard where flagged requests appear as labeled records. The walkthrough targets Node.js 22.12 or later and uses Fastify 5 alongside the @webdecoy/fastify and @webdecoy/node packages.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Cloudflare's AI Bot Rules Silently Block Googlebot, Bypassing robots.txt

A September 2026 Cloudflare policy change altered default settings for AI-related crawlers, causing Googlebot, Bingbot, and Applebot to receive 403 errors on ad-supported pages across new and free-tier accounts. Because Cloudflare classifies Googlebot as a mixed-use crawler serving both search and AI training, it falls under the more restrictive AI training block when that setting is enabled. The problem goes undetected because robots.txt still shows an open allow directive, while the actual block is enforced at the CDN edge layer before the origin server is ever reached. Cloudflare's own data shows Googlebot's crawl share dropped to 27.49% in Q2 2026, down sharply from 57.20% a year prior. Site owners can diagnose the issue by comparing HTTP response codes for browser versus Googlebot user-agent requests, or by using Google Search Console's URL Inspection tool for verified confirmation.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

AfterImage v1.5 syncs live F1 session states with photo archive using a shared clock

Developer building AfterImage, an independent F1 archive using TypeScript, React, Next.js, and Supabase, has detailed the approach taken in version 1.5 to synchronise live race timing with a photo archive. The update introduces a race-weekend hub that separates scheduled session windows from provider-confirmed live states, preventing the interface from falsely marking a session as live due to delays. Browser clients use server timestamps interpolated with performance.now() rather than device clocks, and a polling fallback ensures a slow response cannot roll displayed state backward. Photographs carry optional metadata fields such as capture time and circuit corner, which are never guessed or substituted with upload time. The project also maintains clear attribution by keeping photographer credits and upload ownership as distinct fields in the photo model.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

How to Detect Crawlers on a Static Astro Site Using Cloudflare Workers

Static Astro sites prerender HTML, meaning crawlers can download pages without triggering browser-based analytics scripts, leaving automated traffic largely invisible to site owners. A developer guide explains how to deploy WebDecoy's edge sensor via Cloudflare Workers to observe all incoming requests — including those from bots that never execute JavaScript. The setup requires a custom domain proxied through Cloudflare and does not need changes to Astro's server rendering configuration or build process. Once installed, the Worker intercepts requests at the network edge before they reach the origin, logging detections in the WebDecoy dashboard. Owners can verify the sensor is working by sending a test curl request with a reserved user-agent string and confirming the labeled detection appears in the dashboard.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

HIVE Builds Middleware Layer to Make Existing Software Natively Usable by AI Agents

A project called HIVE is developing an intermediary layer that sits between existing software products and AI agents, enabling direct access to product functionality without screen-scraping or pixel-reading. The platform translates dashboards, forms, workflows, and other product features into structured capabilities that AI agents can understand and execute. HIVE also incorporates a permissions system that lets teams control which product areas an agent can access and which actions require human approval. The team shared this overview on Day 8 of their development log, framing it as a solution to the problem of AI agents being forced to use software not originally designed for them. Further details on how the translation layer works are expected to be published in the next update.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Supabase Gains Ground Over Firebase as Developers Shift to Relational Backends

Firebase has long been the go-to backend-as-a-service for startups and mobile developers due to its ease of setup and real-time document syncing, but its NoSQL architecture creates structural challenges as applications scale. Unlike Firebase's schema-less document model, Supabase is built on PostgreSQL, offering strict data typing, foreign keys, and relational integrity. A key pain point with Firebase is its lack of JOIN support, forcing developers to run multiple sequential queries and manually combine results — a pattern that increases costs and reduces performance. Supabase resolves this by exposing native PostgreSQL through auto-generated REST and GraphQL endpoints, enabling complex relational queries in a single request. The two platforms also differ in security enforcement: Firebase uses a proprietary rules language, while Supabase leverages PostgreSQL's native Row Level Security with standard SQL expressions.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Google Expands Gemini Skills to Workspace in Controlled Beta Rollout

Google is rolling out Gemini Skills to Google Workspace, allowing organizations to create reusable AI instructions built around internal rules, templates, and reference files. The feature is currently in a controlled beta, requiring administrators to enable the Gemini Beta setting in the Admin console before their teams can access it. For individual users, Skills are gradually becoming available in Gemini chats starting with personal Google Accounts, though work and school accounts may not yet have access. Unlike one-off prompts, Skills are designed to package repeatable guidance so multiple team members can apply consistent instructions across Workspace tools like Gmail, Docs, Drive, and Slides. Broader distribution controls and Marketplace integration are listed as coming soon and are not yet available.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Developer open-sources Deiko, a Mac app that gives coding agents voice and visual context

A developer has open-sourced Deiko, a macOS menu-bar app built over two months to help users communicate more effectively with AI coding agents like Claude Code, Cursor, and Codex. The app lets users double-tap a keyboard shortcut to record voice input while pointing at or drawing around screen regions, then automatically compiles a structured brief with cropped screenshots, accessibility labels, and window context. A local pipeline uses on-device models to file each brief into relevant past tasks, giving agents persistent memory across sessions without sending raw screenshots or sensitive data off-device. Deiko also ships a local MCP server so agents can query earlier work, and a Stop hook prompts Claude Code to save outcomes before ending a session. The project is released under the MIT licence on GitHub, with free hosted transcription up to two hours per month and support for on-device or user-supplied API keys.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Developer Builds Vault-Secured Ansible Automation on AWS EC2, Documents Real-World Pitfalls

A developer spent a single day learning and implementing a fully functional Ansible setup on real AWS EC2 instances, complete with Vault-secured secrets and dynamic inventory. The project automates Nginx deployment across multiple EC2 servers without requiring manual access to each machine. Along the way, the developer encountered practical infrastructure challenges, including EC2 public IPs changing on every stop/start cycle. These issues, absent from most tutorials, were highlighted as the true learning moments of working with real cloud infrastructure. The full code and documentation have been published on GitHub for others to reference.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Opensource

Hey everyone! I’m excited to make my first post in the developer community! I’m currently starting my journey into open source and would love to learn from people who have already been through it. I’m especially interested in opportunities like: Google Summer of Code (GSoC) Hacktoberfest Contributing to real-world open-source projects Finding beginner-friendly communities and projects I’m still figuring out how to get started — from choosing projects and making my first contribution to understanding Git/GitHub workflows and writing good PRs. If you have any resources, roadmaps, YouTube channel

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

How to Control Storage Costs in Node.js Notification Services Using Smarter Error Capture

A Node.js notification service using Postgres cron workers can generate excessive storage costs when retry attempts for failed push jobs are each logged as separate failure records. The core issue is that a single failed delivery retried multiple times produces multiple near-identical records, inflating storage bills without adding meaningful business insight. Developers are advised to separate logical job records, individual attempt details, and terminal outcomes into distinct evidence classes with different retention windows. Attempt-level detail should be kept briefly and sampled for successes, while terminal failure records must always be retained and tied to the durable state transition that marks a job as exhausted. Cost attribution should be grouped by tenant and media publication rather than by queue implementation or worker hostname to accurately reflect business ownership.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Security researcher finds 60+ live Supabase keys in public repos, tests if AI can catch them

A security researcher discovered more than 60 live Supabase service_role keys exposed in public GitHub repositories over just three days of scanning, with keys found in config files, .env templates, deploy docs, and even browser-bundled JavaScript. These keys bypass Row Level Security entirely, granting full read/write access to real production databases. The researcher then built a benchmark on Kaggle to test whether nine leading AI models could identify such credential leaks when acting as security reviewers. Results were mixed: Claude Sonnet 5, Gemini 3.7 Flash, Gemini 3 Flash Preview, and Gemini 3.1 Flash Lite each scored a perfect 10 out of 10, while GPT-5.4-nano scored 8/10 and DeepSeek-R1 scored 0/10 due to flagging every case as critical. The findings highlight both the widespread risk of accidental credential exposure and the varying ability of AI coding assistants to detect the very leaks they may have helped create.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Causal Context Cuts Claude AI Agent Costs and Time by Up to 5x in Tests

A experiment by Causely compared two Claude Managed Agents diagnosing injected faults in a 36-microservice Go application running on Kubernetes. Both agents received identical access to cluster data, Grafana metrics, and source code, but one additionally used Causely's MCP server for causal context. The agent with causal context used up to 7 times fewer tool calls, completed tasks up to 5.7 times faster, and cost up to 5 times less than its counterpart. Three fault scenarios were tested, each placing the named service two to three hops away from the actual fault source to simulate real-world incident conditions. Both agents produced correct fixes in every scenario, highlighting that causal context improved efficiency without sacrificing accuracy.

0 comments Read more at DEV Community
← NewerPage 774 of 3167Older →