SShortSingh.
TopTechnologyAI & MLStartupsProgrammingScienceBusinessFinanceCrypto & Web3World

Programming

LatestTopDiscussed
0
ProgrammingDEV Community ·

Developer Proposes Open Standard for AI Agent Identity Lifecycle Semantics

A developer has published a position piece on DEV Community arguing that open standards for AI agent identity must go beyond token formats and API shapes to address lifecycle semantics. The author contends that critical questions — such as how agents are created, what pre-ownership states mean, and how delegation and provenance are tracked — remain unresolved in current approaches. They introduce SAL (available at sal-protocol.dev) as an attempt to define this lifecycle layer, with Vibebase serving as its reference implementation. The piece warns that standardizing too low in the technical stack risks entrenching compatibility around the wrong abstractions, especially as autonomous agents present concerns that traditional user and service authentication models were never designed to handle. The author invites community pushback on what should be in or out of scope for any emerging agent identity standard.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

How Engineers Build a Production-Ready Multi-Agent Banking AI on Google Cloud

A major retail bank handling 400,000 monthly support calls set out to reduce telephony costs by replacing routine customer interactions with an AI-powered banking assistant. Engineers began with a simple LLM-backed chatbot but quickly encountered tool overload when connecting it to 20–35 internal banking APIs, causing unreliable tool selection and high latency. To address this, the system was redesigned as a multi-agent architecture using Google Cloud's Vertex AI Agent Engine, with specialized sub-agents handling accounts, transactions, and service requests under a central orchestration layer. The Model Context Protocol (MCP) was introduced to decouple the agent's reasoning logic from direct API integrations, improving modularity and maintainability. The final production design also incorporates zero-trust authentication, PII masking, and full observability across Google Cloud Platform to meet enterprise security and compliance standards.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Why a Single-Threaded Rust TCP Server Can Outperform Multi-Threaded Designs

A technical deep-dive published on DEV Community argues that single-threaded servers are not an architectural shortcut but a legitimate design choice for network applications. The article explains that in multi-threaded systems, shared state requires synchronisation tools like Mutexes, which can reduce ten concurrent threads to the effective throughput of one. Using Rust as the demonstration language, the author builds a TCP server from scratch to illustrate how a single thread spends most of its time idle while waiting on network I/O rather than doing actual computation. The piece draws an analogy to a railway ticket clerk who works quickly but waits on slow passengers, mirroring how a server thread waits on data travelling across routers and cables. Developers with JavaScript or TypeScript backgrounds are specifically cited as already familiar with this single-threaded, async mental model that Rust can replicate at the systems level.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Docker Networking Explained: Bridge, Host, None, and MacVLAN Modes

Docker offers multiple network types to control how containers communicate with each other and the outside world. The default bridge network connects containers by IP address only, while custom bridge networks enable DNS-based communication, allowing containers to reach each other by name. Host networking removes isolation by sharing the host's IP and port space directly, making it the fastest but most collision-prone option. The 'none' network mode fully isolates a container from all external communication, making it suitable for batch jobs or cron tasks. MacVLAN assigns a real IP and MAC address to a container on the physical network, useful for integrating with legacy systems that require direct Layer 2 connectivity.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Magento 2.4.6 Lost Security Support in August, Leaving Stores Exposed

Magento Open Source 2.4.6 stopped receiving security patches on 11 August 2026, while Adobe Commerce customers get one additional year of coverage. The underlying PHP versions — 8.1 and 8.2 — have also lost or are losing security support, compounding the risk even for paid-licence holders. Recent vulnerabilities illustrate the danger: the CosmicSting flaw in 2024 led to payment skimmers on over 4,000 stores, and in 2025 some 62% of Magento sites remained unpatched six weeks after an emergency fix for a critical remote code execution bug. Stores still on 2.4.6 cannot meet PCI DSS patch requirements in the standard way, as no new fixes will be issued for the version. Upgrading to a supported release requires updating the hosting stack and auditing third-party extensions, but security experts warn that delaying further increases exposure with no remediation path available.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Dicionário

Palavras Difíceis Redes de computadores https://www.ibm.com/br-pt/think/topics/networking APi Porque API REST? Servidor de aplicação https://www.ibm.com/br-pt/think/topics/application-server JWT A partir daí, você não precisa mostrar sua identidade toda vez que for pedir uma bebida ou entrar No mundo digital, o JWT é essa pulseira. Ele é um passe seguro que viaja junto com os seus Keycloak Redis https://redis.io/docs/latest/develop/clients/nodejs/ MockServer https://www.mock-server.com/ Load Balancer (balanceador de carga) https://www.vmware.com/topics/load-balancing Bibliotecas (Lib) Uma bibl

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

AI Writes 42% of GitHub Code in 2026, But Bugs and Gaps Keep Developers Needed

By 2026, AI tools like GitHub Copilot and GPT-5 Turbo generate roughly 42% of code on GitHub and cost as little as $10 per month, prompting many companies to reduce developer headcount. However, a 2026 Snyk study found that 70% of AI-generated code samples contained critical security vulnerabilities, and Shopify reported that 18% of AI-assisted pull requests required major rewrites due to missed business logic. Teams that rely solely on AI have been found to spend 2.4 times more time in debugging sprints, according to the 2026 Stack Overflow survey, making the apparent cost savings a false economy. McKinsey data from the same year shows that 68% of failed AI-only projects lacked domain expert oversight, highlighting that human judgment remains essential for requirements, security review, and product decisions. Developers in 2026 are increasingly shifting from writing raw code to roles focused on system design, prompt engineering, and validating AI output.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Picsart Integrates Google Gemini for AI-Powered Logo and Image Workflows

Picsart has integrated Google's Gemini 3.7 Flash model into its API platform to support AI-driven logo generation and image editing workflows. The Logo Generator endpoint accepts structured brand inputs — including brand name, business description, color tone, and logo description — to help users move from a creative brief to early visual concepts more efficiently. According to Picsart's documentation and a Google Cloud case study, the company uses the Gemini Enterprise Agent Platform to access frontier multimodal models and refine prompts in production. The integration is aimed at developers, marketing teams, and agencies looking to build repeatable, API-based branded-content workflows within a single platform. No general public rollout date or pricing details have been confirmed, and Picsart advises interested businesses to verify access terms and costs directly before building production processes around the capability.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Atlassian's Jira AI Agents Could Shift Developers Into Reviewer and Architect Roles

Atlassian has introduced AI Agent capabilities for Jira that can autonomously pick up backlog tickets, implement code changes, run tests, and raise pull requests for human review. This mirrors a broader shift already underway in software development, where engineers increasingly review AI-generated code rather than writing every line themselves. However, the approach has clear limits — complex features, ambiguous requirements, and architectural decisions still require significant human expertise. The developer role may consequently evolve through stages: from writing code, to writing prompts, to reviewing outputs, and ultimately to defining problems and designing solutions. The emerging view is that the most valuable engineering skill may no longer be coding speed, but the ability to clearly frame problems and verify that AI-built solutions are correct.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Claude Code Billing Explained: Four Token Types and Cache Timing Drive Your Costs

Claude Code charges developers across four token categories — input, output, cache write, and cache read — each priced differently, with output tokens costing roughly five times more than input tokens. Cache read tokens are the cheapest at one-tenth the base input rate, making prompt caching essential for cost efficiency on large codebases. Two cache durations exist: a 5-minute TTL for API key users and a 1-hour TTL for Pro/Max subscribers, with the longer TTL costing 60% more to write but offering greater savings for sessions with natural pauses. Allowing the cache to expire mid-session triggers a full rewrite that can cost up to 20 times more than a simple cache hit, and actions like switching models, running /compact, or toggling plugins also reset the cache. Developers can reduce unexpected costs by enabling finish notifications and monitoring the live status line to stay within their cache window.

0 comments Read more at DEV Community
0
ProgrammingHacker News ·

Guide: Configuring Swap, ZRAM, Zswap and Hibernate on NixOS

A technical blog post by Matthew Brunelle walks through memory management options available on NixOS, covering swap, ZRAM, Zswap, and hibernation. The guide explains the differences between these approaches and how to configure them within the NixOS ecosystem. ZRAM and Zswap are compressed memory techniques that can improve system performance under memory pressure. The article serves as a practical reference for NixOS users looking to optimize their system's memory and power management settings.

0 comments Read more at Hacker News
0
ProgrammingHacker News ·

Anthropic's Claude AI Identifies Novel Enzyme System Featuring CRISPR-Like Repeats

Anthropic's AI system Claude has reportedly discovered a previously unknown enzyme system that contains repeating sequences similar to those found in CRISPR. The finding was announced by Anthropic and represents an application of large language models to biological research. The discovery suggests AI tools may be capable of identifying complex molecular patterns that could have implications for gene-editing and related fields. Details of the enzyme system and its potential applications were shared via Anthropic's official news channel. The development has drawn attention from the research and technology communities, sparking discussion about AI's growing role in scientific discovery.

0 comments Read more at Hacker News
0
ProgrammingHacker News ·

Microsoft Blog Traces the Forgotten History of Windows Scroll Bar Shortcuts

A post on Microsoft's Old New Thing developer blog explores the historical evolution of scroll bar keyboard shortcuts in Windows. The article examines how these shortcuts developed over time and the design decisions behind them. Published in September 2026, the piece attracted attention on Hacker News, garnering 14 points and community discussion. The blog, maintained by longtime Microsoft engineer Raymond Chen, is known for documenting obscure Windows internals and legacy design choices.

0 comments Read more at Hacker News
0
ProgrammingDEV Community ·

Codename One Replaces Wrapped Javadoc with Hugo-Integrated API Docs

Codename One, an open-source framework for building cross-platform apps from Java or Kotlin, has overhauled how its API documentation integrates with its developer website. Previously, Javadoc pages were loaded through a JavaScript wrapper, causing mismatches in navigation, styling, and dark mode between the API docs and the rest of the site. The team built a new Maven doclet that feeds the API model directly into their Hugo static site generator, allowing both tools to handle what they do best. Each Java type now generates a Hugo content file rendered with the same theme and typography as the guide and blog, while a standard Javadoc archive is still produced for offline use. A redesigned API search index, compressed to 341 KB gzipped, lets users find API members alongside other site content within a unified interface.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Build vs Buy: Why Manufacturers Should Rethink CRM Strategy for External Users

Most manufacturers default to buying off-the-shelf CRM software, but the real cost and complexity lie in the external-facing layer used by dealers, distributors, and service technicians. Standard CRM platforms are built around contacts and pipelines, while manufacturing sales depend on ERP-held data like contract pricing, part codes, and production slots. Enterprise options such as Salesforce Manufacturing Cloud and HubSpot Sales Hub carry steep per-seat fees, and external portal users — often outnumbering internal staff — can nearly double the annual licence bill. A hypothetical mid-size manufacturer with 15 internal users and 300 dealer accounts could face over $139,000 per year in list-price Salesforce licences alone. This cost structure, combined with quoting complexity and the recent end-of-sale of Salesforce CPQ, is prompting manufacturers to evaluate which CRM components are worth building in-house versus purchasing.

0 comments Read more at DEV Community
0
ProgrammingHacker News ·

UK Military Confirms It Jams Foreign Satellites as Defensive Measure

The UK military has disclosed that it is actively jamming satellites operated by other nations as part of its defensive strategy. This revelation was made to the BBC, marking a rare public acknowledgment of such electronic warfare capabilities. The practice involves disrupting or degrading satellite signals to protect British interests and operations. The disclosure highlights the growing role of space-based electronic warfare in modern military doctrine. Details about the specific targets or scale of the jamming operations were not publicly revealed.

0 comments Read more at Hacker News
0
ProgrammingDEV Community ·

Mapping Legacy Position Codes to Oracle HCM: Why a Translation Layer Matters

Data migration to Oracle HCM often hits a critical snag when source systems store worker assignments using a single legacy position code, while Oracle requires multiple distinct values such as job code, department, and business unit. Organizations frequently redesign their job and position structures during implementation, meaning legacy codes rarely map directly to Oracle values. A mapping-table approach — rather than hardcoded IF-THEN logic — serves as a translation layer, making conversions easier to maintain, test, and update as Oracle codes evolve. This strategy also improves governance by centralizing all mappings for review by functional and business teams. Crucially, worker assignments can only be loaded after departments, jobs, and positions have already been created in Oracle HCM, a dependency chain that project teams often underestimate during planning.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

How Tekton, Cosign, and Kyverno Create a Fully Signed Kubernetes CI/CD Pipeline

A homelab engineer has documented a four-component CI/CD security stack built on RKE2 Kubernetes, ensuring every container image is built, scanned, signed, and verified before deployment. The pipeline uses Tekton for in-cluster image builds, SonarQube and Trivy for code quality and vulnerability scanning, and Tekton Chains to attach cryptographic signatures and SLSA provenance to each artifact. Images are stored in a self-hosted Harbor registry alongside their signatures, provenance data, and software bill of materials. Kyverno acts as the admission gatekeeper, rejecting any image at the cluster door that does not carry a valid cryptographic signature. The setup replaces Jenkins and Nexus with cloud-native alternatives, reducing complexity while enforcing an end-to-end verifiable software supply chain.

0 comments Read more at DEV Community
← NewerPage 983 of 3190Older →