Articles from Source: The-New-Stack

Next.js in ChatGPT: Vercel Brings the Dynamic Web to AI Chat

2025-10-30 14:00
🚀 Vercel’s Andrew Qu discusses the challenges of integrating Next.js with OpenAI's ChatGPT Apps platform. Initially, ChatGPT apps were designed as static HTML pages, limiting their dynamic capabilities. While they can respond to user input, they operate within strict security constraints. Vercel has found ways to enhance functionality, allowing Next.js to run in ChatGPT's environment without altering the framework itself. This opens new possibilities for dynamic web experiences in AI chat!...
Source: The New Stack
Richard MacManus

Addressing 3 Failure Points of Multiregion Incident Response

2025-10-30 12:00
🌍 As organizations expand globally, managing incidents across multiple cloud regions becomes more complex. Key challenges include fragmented tools, context loss during handoffs, and increased debugging needs. Each region often uses different systems, leading to confusion and inefficiencies. To address these issues, standardizing tools and processes is essential. This ensures that all teams can respond effectively, regardless of location. Adopting AI-assisted scheduling can also help optimize...
Source: The New Stack
Cristina Dias

Eclipse Opens Up Enterprise AI Agent Development With ADL

2025-10-29 21:00
🚀 The Eclipse Foundation has introduced the Agent Definition Language (ADL), aiming to challenge proprietary AI platforms. ADL is designed to enhance enterprise AI development, making it more accessible and scalable. It is the core of the Eclipse LMOS, an open-source platform for building AI agents. Project lead Arun Joseph developed this tool to overcome challenges faced during his work at Deutsche Telekom, where existing infrastructure hindered the adoption of new frameworks. Eclipse's...
Source: The New Stack
Darryl K. Taft

Building Multiagent Workflows With Microsoft AutoGen

2025-10-29 20:00
Explore the capabilities of Microsoft AutoGen in building multiagent workflows! 🌐 Traditional AI often feels limited, but AutoGen allows multiple AI agents to collaborate, enhancing problem-solving. By assigning roles like planning, generating, and critiquing, these agents can effectively tackle complex tasks together. Real-world applications, such as research analysis and coding, showcase this innovative approach. Learn how to implement AutoGen with a simple tutorial to elevate your AI...
Source: The New Stack
Oladimeji Sowole

How To Build a Serverless API With Bun and Hono

2025-10-29 19:01
🚀 Discover how to build a serverless API using Bun and Hono! Bun offers a modern alternative to Node.js, focusing on speed and simplicity. Hono, a lightweight web framework, complements Bun by making development efficient across various JavaScript runtimes. Explore the benefits of this combination, including fast API responses and TypeScript compatibility. Ready to get started? Check out the full tutorial! #ServerlessAPI #Bun #Hono #WebDevelopment #JavaScript
Source: The New Stack
Jessica Wachtel

What Good Software Supply Chain Security Looks Like

2025-10-29 18:00
Organizations using open source software are increasingly facing complex security and compliance challenges. A recent report highlights a 156% rise in malicious packages in 2024. To enhance software supply chain security, teams must focus on reducing vulnerabilities and removing unnecessary code from container images. Distroless images can help by minimizing attack points, making applications safer. #SoftwareSecurity #OpenSource #CyberSecurity #DevOps #SupplyChain
Source: The New Stack
Rita Manachi

How Are OpenTelemetry and Fluent Bit Related?

2025-10-29 17:00
Explore the connection between OpenTelemetry (OTel) and Fluent Bit in the context of Kubernetes logging! 📊 Fluent Bit captures both logs and local metrics, enhancing operational monitoring. Its alignment with OTel standards allows seamless integration without major changes. For deeper insights, check out the full book on Fluent Bit and its evolving role in observability. 📚 #OpenTelemetry #FluentBit #Kubernetes #Observability #TechInsights
Source: The New Stack
Phil Wilkins

New Open Source Tool from Angular Scores Vibe Code Quality

2025-10-29 16:00
The Angular team at Google has launched an open-source tool called Web Codegen Scorer. This tool evaluates code generated by large language models (LLMs) for Angular applications. It focuses on best practices, accessibility, and security in frontend development. The team aims to address challenges in relying on AI-generated code, which may not always be production-ready. Web Codegen Scorer serves as a fitness test for LLM-generated code, helping developers ensure quality. #Angular #OpenSource...
Source: The New Stack
Loraine Lawson

Confluent Provides a Real-Time Context Engine for AI

2025-10-29 15:00
🚀 Confluent Inc. aims to enhance AI capabilities by offering a real-time context engine, ensuring timely AI analysis and decision-making. Their platform leverages the open-source Kafka and Flink for effective data streaming. As organizations invest heavily in generative AI, Confluent emphasizes the importance of continuous data flow for valuable insights. At the Current conference, they introduced Confluent Intelligence, which includes tools for optimizing data for AI operations. This...
Source: The New Stack
Joab Jackson

Shadow AI: The Growing Risk IT Leaders Must Address

2025-10-29 14:30
🚨 The rise of Shadow AI presents new challenges for IT leaders. As employees increasingly use third-party AI tools without approval, risks to cybersecurity grow. IT departments strive to balance providing innovative solutions while maintaining security. To learn more about managing these risks, check out Anaconda's 2025 "State of Data Science and AI Report." 📈 #ShadowAI #Cybersecurity #DataScience #ITLeadership #AITrends
Source: The New Stack
Steve Croce

dbt Labs Open Sources MetricFlow: An Independent Schema for Data Interoperability

2025-10-29 13:30
🚀 dbt Labs has open-sourced MetricFlow, a SQL generation tool that strengthens the dbt semantic layer under the Apache 2.0 license. This release supports the Open Semantic Interchange (OSI) Initiative, aiming for data standards across platforms. It includes a JSON-based metadata layer, providing a universal schema for data understanding. This could enhance interoperability between data systems and improve transparency in statistical AI applications. 🛠️📊 #DataInteroperability #OpenSource...
Source: The New Stack
Jelani Harper

LogicStar AI Wants To Automate Debugging

2025-10-29 12:00
🚀 LogicStar AI is set to revolutionize software debugging! Founded by Boris Paskalev, LogicStar aims to automate the entire process, from identifying bugs to fixing them without human intervention. The platform boasts a 90% accuracy rate in bug resolution, allowing developers to focus on other tasks. LogicStar analyzes code deeply during onboarding, ensuring effective bug detection and validation. This innovative approach could streamline release cycles significantly! #LogicStarAI...
Source: The New Stack
Meredith Shubel

How MCP and AI are Modernizing Legacy Systems

2025-10-28 20:03
Large enterprises often face a divide in their IT landscape, balancing modern cloud-native applications with rigid legacy systems. A new strategy uses agentic AI and the Model Context Protocol (MCP) to build an intelligent layer over these legacy systems instead of replacing them. This allows for better integration without the risks of a complete overhaul. However, it requires a cultural shift in DevOps practices to address the blind spots around legacy systems. Insights from industry leaders...
Source: The New Stack
Saqib Jan

Fast, Rust-Based Zed Code Editor Finally Arrives on Windows

2025-10-28 16:58
🚀 Zed, the Rust-based code editor from the creators of Atom and Tree-sitter, has launched on Windows! This fast editor offers GPU acceleration and aims to support various programming languages. While there are minor concerns, the team shows commitment to improving the Windows experience. Zed is positioned as a useful tool alongside major editors like Visual Studio and Sublime Text. 🔧 #CodeEditor #Zed #WindowsDevelopment #ProgrammingTools #RustLang
Source: The New Stack
David Eastman

GitHub Embraces the Coding Agent Competition With Agent HQ

2025-10-28 16:10
🚀 GitHub has unveiled Agent HQ at the Universe 2025 conference, introducing third-party coding agents to the platform. Developers will gain visibility into their agents' processes and have the ability to manage code generation from one location. This feature will support notable agents like Claude Code and Codex later this year. Additionally, Copilot Pro+ users can now integrate OpenAI’s Codex into VS Code, enhancing the coding experience. GitHub aims to empower developers with customizable...
Source: The New Stack
Frederic Lardinois

The Great AWS Outage: The $11 Billion Argument for Kubernetes

2025-10-28 16:00
The recent AWS outage highlighted vulnerabilities in cloud infrastructure. For eight hours, major services, including e-commerce and healthcare, were disrupted, resulting in an estimated loss of over $11 billion. 💸 This incident raises questions about reliance on single-cloud providers. Many companies built systems around a sole provider, exposing them to significant risks. As the industry shifts towards "multicloud" strategies, it's essential to recognize the complexities involved. True...
Source: The New Stack
Arjun Iyer

Why Sudo-rs Brings Modern Memory Safety to Ubuntu 26.04

2025-10-28 14:30
🚀 Ubuntu 26.04 will introduce sudo-rs, a version of sudo written in memory-safe Rust. It will coexist with the traditional C-based sudo, ensuring users can choose their preferred version. This shift aims to enhance security, maintainability, and flexibility in managing Linux privileges. Marc Schoolderman highlights that this effort is part of a broader initiative to rewrite critical utilities in memory-safe languages. Support for the project and its maintainer, Todd Miller, is essential....
Source: The New Stack
Steven J. Vaughan-Nichols

More Than a Warm-Up: Start KubeCon With a Colocated Event

2025-10-28 14:00
KubeCon + CloudNativeCon is evolving beyond keynotes and pavilions. Colocated events are becoming essential for attendees, offering targeted learning opportunities before the main conference. From newcomers to seasoned pros, there’s something for everyone. 🌟 One highlight is KyvernoCon, focusing on Policy as Code, which aids enterprises in managing Kubernetes efficiently. 🚀 GitOps is also gaining traction with the introduction of FluxCon, emphasizing the growing importance of these practices...
Source: The New Stack
Cortney Nickerson

Beyond ‘Shift Left’: Improving AI Training Data

2025-10-28 13:00
The software development industry faces an "engineering productivity paradox." While AI coding assistants generate a significant amount of code, productivity gains are limited to around 10%. This issue arises from the need for human developers to review and fix AI-generated code, which often contains flaws from the unfiltered data it learns from. Studies highlight that leading LLMs produce code with high-severity vulnerabilities, creating a burden on human reviewers. To address this, the...
Source: The New Stack
Manish Kapur

How Anyshift Builds the Context for Faster, Smarter Incident Response

2025-10-28 12:00
🌐 Modern companies face challenges with complex, multicloud architectures, impacting incident response times for site reliability engineers (SREs). 🚨 A 2023 survey revealed that 62% of IT professionals reported longer resolution times due to fragmented infrastructures. 🔍 Anyshift, co-founded by Roxane Fischer, aims to address these issues by providing context through a continuously updated infrastructure graph and a smart assistant named Annie. #IncidentResponse #SRE #TechInnovation #Anyshift...
Source: The New Stack
Meredith Shubel

What Does It Take for AI Agents To Deploy Infrastructure?

2025-10-27 18:00
AI is transforming the coding landscape, but deploying infrastructure remains a challenge. Many organizations still depend on ticket queues and manual processes, causing delays. Although AI can generate code, the complexity of infrastructure management poses significant barriers. Key challenges include: 🔍 Lack of organizational context 🛠️ Complex tech stacks with hidden dependencies ⚖️ Risks related to compliance and cost management Understanding these barriers is essential for effective AI...
Source: The New Stack
Idan Yalovich

What Are AI Orchestrators and Why Do They Matter Now?

2025-10-27 17:00
AI orchestrators play a crucial role in coordinating AI agents, ensuring they don't operate in isolation. They manage components, processes, and information exchange for efficient performance. Zachary Hanif from Twilio emphasizes that understanding these tools helps developers optimize AI workloads and set appropriate guardrails. As AI security concerns grow, orchestrators also help prevent potential threats. This coordination is likened to the well-organized processes in a McDonald’s...
Source: The New Stack
Agam Shah

Why Your Platform Engineering Career Is Really a Sales Job

2025-10-27 16:04
Platform engineering is more than just tech—it's about effectively communicating value to various stakeholders. At a recent conference, Christopher Batey emphasized the need for platform engineers to adopt a sales mindset. Engaging with senior leadership, product management, and security is crucial for demonstrating the impact of internal developer platforms (IDPs). With 45% of platform teams disbanding within 18 months, understanding and explaining the "why" is essential for success. 💡...
Source: The New Stack
Jennifer Riggins

GitHub Bets on Openness

2025-10-27 15:03
GitHub is navigating changes following CEO Thomas Dohmke's departure and its deeper integration with Microsoft. In a recent interview, COO Kyle Daigle emphasized that GitHub's core mission remains focused on supporting developers and open-source projects. He reassured the community that the platform's identity will not change, despite concerns about Microsoft's involvement. Daigle highlighted GitHub's unique position in the developer ecosystem and its commitment to prioritizing developers'...
Source: The New Stack
Frederic Lardinois

What’s Next for Developer Teams: How To Prepare Now

2025-10-27 14:00
🌐 The article discusses the evolving landscape for developer teams in the AI era. As AI capabilities increase, technology leaders are questioning how to foster critical-thinking skills that AI cannot replicate. Despite fears of AI replacing developers, research shows 95% of enterprise AI pilots fail to deliver value. The piece highlights the need for new roles: "cognitive architects" and "AI guardians." These positions will focus on system-level thinking and ensuring AI-generated applications...
Source: The New Stack
Emilio Salvador

Cloud Native and AI: Why Open Source Needs Standards Like MCP

2025-10-27 13:00
The article discusses the importance of open standards like the Model Context Protocol (MCP) as AI and cloud native technologies converge. 🌐 Recent contributions, such as the MCP server for Argo CD, highlight how these standards facilitate integration between AI systems and infrastructure tools. This will streamline operations and reduce manual efforts. 🤖 MCP aims to unify AI interactions with various cloud native environments, fostering collaboration and innovation within the community. 🚀...
Source: The New Stack
Alexander Matyushentsev

Valkey 9.0 Debuts Multidatabase Clustering for Massive-Scale Workloads

2025-10-27 12:00
🚀 Valkey has launched version 9.0, its most significant update since its inception. This release introduces multidatabase clustering, allowing developers to manage separate databases in distributed clusters without sacrificing performance. 🔄 Key features include atomic slot migration for seamless rebalancing and enhanced reliability, especially in enterprise and cloud settings. 🔒 Security enhancements address vulnerabilities in Lua scripting, improving client authentication options. This...
Source: The New Stack
Steven J. Vaughan-Nichols

OpenTelemetry Adoption Update: Rust, Prometheus and Other Speed Bumps

2025-10-26 18:00
OpenTelemetry is gaining traction as a standard for observability in organizations. It allows for standardizing telemetry data across various platforms. 📊 Currently, support for Rust is in beta, with ongoing development for traces, metrics, and logs. Other major programming languages have more established support. Implementation remains complex, especially in large environments, posing challenges for widespread adoption. 🛠️ #OpenTelemetry #Observability #Rust #TechUpdate #CNCF
Source: The New Stack
B. Cameron Gain

The AI Inflection Point Isn’t in the Cloud, It’s at the Edge

2025-10-26 15:00
AI model development is shifting from cloud to edge devices, offering high-performance computing capabilities closer to data sources. Zach Shelby, CEO of Edge Impulse, highlighted the challenges of scalability and power in deploying generative AI. Edge devices can run various language models, but large language models (LLMs) face significant hurdles. Understanding the value of LLMs in industrial settings remains a challenge, as companies seek clear returns on investment. Meanwhile, visual...
Source: The New Stack
Alex Williams

Ubuntu 25.10 Scraps X11 for Wayland: A Solid Step Forward

2025-10-26 14:00
🚀 Ubuntu 25.10 marks a significant shift by replacing X11 with the Wayland protocol for its GNOME desktop. This change aims to enhance performance but may pose compatibility issues for some legacy Linux applications. Key updates also include GNOME 49, new default apps, improved security features, and kernel upgrades. Explore the latest enhancements and consider how they may impact your Linux experience! 🖥️🔒 #Ubuntu #Linux #Wayland #GNOME #OpenSource
Source: The New Stack
Jack Wallen

Ken Thompson Recalls Unix’s Rowdy, Lock-Picking Origins

2025-10-26 13:05
Ken Thompson reflects on the early days of the Unix operating system in a recent oral history at the Computer History Museum. 🎤 At 82, he shares memories of the collaborative environment at Bell Labs, where innovation thrived among a spirited group of tech enthusiasts. 💡 Thompson details his transition from the failed Multics project to the creation of Unix, emphasizing the importance of experimentation and community. He also shares a quirky anecdote about raising a baby alligator at the lab!...
Source: The New Stack
David Cassel

Why PyTorch Won

2025-10-25 20:00
During the PyTorch Conference 2025 in San Francisco, I had the opportunity to interview Luca Antiga, a key figure in the PyTorch Foundation. Antiga shared insights on why PyTorch has thrived in the AI landscape. Its user-friendly, "pythonic" nature has made it a favorite among researchers, who have since transitioned to the tech industry. Moreover, PyTorch has maintained relevance through various AI advancements, including generative AI. It’s now vital for both model training and production,...
Source: The New Stack
Frederic Lardinois

AWS Adds Scope 3 Emissions Data to Customer Carbon Footprint Tool

2025-10-25 19:00
📢 AWS has updated its Customer Carbon Footprint Tool (CCFT) by adding Scope 3 emissions data for the first time. This update, based on an enhanced methodology, allows for better year-over-year comparisons. The historical data has been recalculated back to January 2022 to support trend analysis. AWS emphasizes data quality and transparency for effective sustainability reporting. The updated CCFT data export now includes emissions breakdowns for Scopes 1, 2, and 3, enhancing customers'...
Source: The New Stack
Charles Humble

Vercel Marketplace Offers Agentic AI Building Blocks

2025-10-25 17:00
🚀 Vercel has launched Agents and Services in its marketplace, allowing developers to integrate AI workflows into projects effortlessly. These tools simplify the integration process by offering unified billing and observability functions, addressing past challenges with separate systems. Two main components are available: 1️⃣ Agents for specialized tasks. 2️⃣ Services for custom agent infrastructure. Notable agents include CodeRabbit for code reviews, Corridor for security monitoring, and...
Source: The New Stack
Loraine Lawson

How to Manage the Growing AI Sprawl in Your SaaS Stack

2025-10-25 14:00
The enterprise software landscape is evolving as AI becomes integral to SaaS platforms. 🤖 Organizations now face the challenge of AI sprawl, where each tool in their stack incorporates its own AI capabilities. This complexity increases management difficulties for engineering teams. Moreover, the costs of AI-enhanced features can significantly raise subscription prices, doubling or tripling overall SaaS expenses. 📈 Integration challenges arise as traditional software architectures struggle to...
Source: The New Stack
Daniel Clydesdale-Cotter

Ditch Python: 5 JavaScript Libraries for Machine Learning

2025-10-25 13:00
🚀 Python has been the go-to language for machine learning, but JavaScript is stepping up. At the devmio International JavaScript Conference, Laurie Lay showcased five open-source libraries aimed at making machine learning accessible to the JavaScript community. Key libraries include Danfo.js for data manipulation and The Natural Library for natural language processing. These tools empower developers to explore machine learning directly within JavaScript. #JavaScript #MachineLearning...
Source: The New Stack
Loraine Lawson

IBM’s Mellea Tackles Open Source AI’s Hidden Weakness

2025-10-24 22:00
IBM's Nathan Fulton highlights a key issue in open source AI: infrastructure disparity. Despite open source models offering advanced capabilities, many companies mistakenly compare them to polished proprietary platforms like OpenAI. This creates an unfair evaluation landscape. To address this, IBM introduced Mellea, a new open source library aimed at leveling the playing field. Fulton emphasizes that while open models may outperform expectations, the lack of integrated software stacks hinders...
Source: The New Stack
Darryl K. Taft

2 Fixes Vastly Cut TiKV Write Stalls From SST File Ingestion

2025-10-24 18:00
🚀 TiKV, the open-source key-value database, has addressed write stalls during SST file ingestion. 🔍 The issue arose when SST ingestion blocked foreground writes to maintain sequence order, leading to latency spikes. 💡 Two enhancements were implemented: reducing unnecessary flushes and allowing concurrent writes during ingestion with safety mechanisms. These changes significantly improve performance and reduce stall times. #TiKV #Database #OpenSource #Performance #TechNews
Source: The New Stack
Jinpeng Zhang

How TARmageddon Compromises Rust Security: A Developer’s Guide

2025-10-24 17:00
🚨 A new Rust vulnerability, TARmageddon (CVE-2025-62518), has been identified by Edera, affecting the tokio-tar library. This flaw may lead to remote code execution (RCE) in several software programs, including popular tools like Astral’s uv and wasmCloud. With a CVSS score of 8.1, it's classified as high severity. The issue arises from a desynchronization bug in TAR parsing that allows attackers to manipulate nested TAR archives. Developers are advised to be cautious, as this highlights...
Source: The New Stack
Steven J. Vaughan-Nichols

How To Build Resilient IT Operations in 4 Steps

2025-10-24 16:00
🔧 Managing digital incidents is now essential for IT operations. Research shows 84% of businesses faced increased outages recently. Operational resilience is key to thriving amidst disruptions. This involves predicting, withstanding, and recovering from IT outages. Here are four steps to enhance resilience: 1️⃣ Assess current operations to identify inefficiencies and outdated systems. 2️⃣ Automate repetitive tasks to streamline workflows. For more insights, read the full article....
Source: The New Stack
Ariel Russo