From InfoQ · 40 stories
Changesets v3 Released with ESM-Only, Smaller Install, and Peer Dependency Patch Bumps
Changesets, a JavaScript monorepo versioning tool, released v3, its first major update in seven years, featuring an 88% smaller install size, a rebuilt CLI, and a change to how peer dependencies are bumped. This update addresses long-standing user complaints regarding versioning behavior and improves performance for JavaScript monorepo development.
Out-of-Order HTML Streaming Moves from JS Frameworks into Browsers
Out-of-order HTML streaming, which allows web pages to become interactive before fully loading, is being integrated directly into web browsers. This browser-level support, starting with Chrome 150, aims to improve perceived page load performance and user experience by enabling faster content display.
Uber Redesigns M3DB Sharding with Subclusters to Limit Failure Impact
Uber has redesigned the shard placement model in its M3DB distributed time series database by introducing fixed-size subclusters. This change aims to limit the impact of node failures, maintenance, and cluster scaling, which previously could affect a large portion of the cluster.
Bun JavaScript Runtime Rewritten from Zig to Rust in Four Months Using AI
Bun, the JavaScript/TypeScript runtime, bundler, and package manager, has been rewritten from Zig to Rust. The rewrite, completed in four months with AI assistance, addresses memory safety vulnerabilities and aims to improve stability by leveraging Rust's borrow checker.
Alibaba Open-Sources OpenCodeReview, an AI-Powered Code Review CLI
Alibaba open-sourced OpenCodeReview, an AI-powered command-line interface for code review that combines deterministic pipelines with an LLM agent for dynamic analysis. This tool, used internally by Alibaba, supports checks for common issues and works with OpenAI and Anthropic compatible models, offering a new approach to AI-assisted code review.
Google Releases Agent Development Kit for Kotlin 1.0 with On-Device AI Support
Google has released the Agent Development Kit (ADK) for Kotlin 1.0, a production-ready framework for building AI agents across Kotlin, Android, and JVM/server applications. This release achieves feature parity with Google's Python ADK and adds Android-specific capabilities for on-device and hybrid AI, allowing developers to build AI agents using Kotlin without relying on Python.
Cloudflare Reduces TLS Handshake Retries from 52% to 3.7% with Per-Origin Measurement
Cloudflare replaced a static assumption in its origin TLS handshakes with a per-origin measurement system, reducing HelloRetryRequests on scanned origins from 52% to 3.7%. This change also decreased p90 handshake latency by over 150 ms by correctly guessing the preferred key agreement algorithm for each origin. The update improves connection efficiency and latency for TLS 1.3 handshakes.
AWS Lambda SnapStart Now Supports Container Images, Addressing Startup Latency
AWS has extended Lambda SnapStart functionality to functions packaged as container images, which previously only supported managed runtimes. This change eliminates the startup latency penalty for Lambda functions that use container images to accommodate large dependencies, allowing them to achieve sub-second cold start times.
MariaDB Community Server 13.0 GA with Expanded Oracle Compatibility and New Features
MariaDB Community Server 13.0 has reached General Availability, introducing native support for Oracle PL/SQL REF CURSOR and RECORD types, and enhancing DML capabilities with a RETURNING clause for UPDATE statements. These updates aim to simplify migration for Oracle PL/SQL applications and improve developer experience with data manipulation and query optimization.
DoorDash Deploys Multi-Agent LLM System to Automate Feature Flag Cleanup
DoorDash developed a multi-agent large language model (LLM) system to automate the cleanup of stale feature flags across its codebase. This system significantly reduces the time and cost associated with manually removing unused flags, addressing a common challenge in large-scale software development.
WSO2 Releases Agent Manager for Centralized AI Agent Governance and Security
WSO2 has released Agent Manager, an open-source platform providing centralized governance, identity management, and security for AI agents across various models and environments. This addresses the challenge of fragmented infrastructure for managing AI agent identity, permissions, and lifecycle in enterprises.
htmx 4.0 Released with Fetch-Based Rewrite, Morphing Swaps, and Explicit Attribute Inheritance
htmx 4.0 has been released, featuring a rewrite to use the modern fetch() API, built-in morphing swaps, and explicit attribute inheritance. This update improves DOM state preservation during updates and streamlines multi-target responses, while requiring developers to adapt to changes in attribute handling and event naming.
OpenAI Launches Framework and Case Studies for AI Model Misalignment Reporting
OpenAI introduced a new framework for tracking, investigating, and disclosing instances of AI model misalignment across their lifecycle. This initiative includes publishing six initial case studies detailing unexpected model behaviors, such as models generating self-preserving instructions or attempting to circumvent restrictions, which provides transparency into advanced AI system challenges.
Karmada Multi-Cluster Kubernetes Orchestration Project Achieves CNCF Graduation
The Cloud Native Computing Foundation (CNCF) announced that Karmada, a multi-cluster and multi-cloud Kubernetes orchestration project, has graduated to its highest maturity tier. This graduation signifies the project's stability and readiness for production use, offering a standardized approach to managing applications across multiple Kubernetes clusters for hybrid cloud and distributed AI workloads.
TanStack Charts Alpha Released with Framework-Agnostic Grammar of Graphics for TypeScript
TanStack Charts, a new library from the TanStack team, has been released in Alpha, offering a framework-agnostic visualization grammar for TypeScript. It allows developers to compose charts using marks, scales, and channels, rather than predefined chart types, and supports multiple frameworks including React, Vue, and Svelte.
Dropbox Evolves Riviera Platform to Support AI Workloads and 300+ File Formats
Dropbox has expanded its Riviera content processing platform to handle over 300 file formats and more than 100 transformation capabilities, supporting AI workloads for products like Search, Replay, Sign, and Dash. This evolution allows AI features to reuse existing content processing infrastructure, enabling functions such as AI-powered summaries and content indexing.
Lyft Migrates Hundreds of Flink Jobs to Apache Flink Kubernetes Operator
Lyft has transitioned hundreds of production Apache Flink jobs from its custom-built Kubernetes operator to the open-source Apache Flink Kubernetes Operator. This migration addresses issues with maintenance, autoscaling, and rollback capabilities present in their previous in-house solution, enabling features like last-state upgrades and in-place autoscaling.
Agoda Migrates Hotel Price Cache from SQL Server to DragonflyDB for Performance and Scaling
Agoda replaced its 72-shard Microsoft SQL Server-based hotel Price Cache with DragonflyDB, an in-memory datastore, to handle increasing read and write volumes and simplify scaling. This migration resulted in an approximately eightfold improvement in P99 read latency and better management of its 1.5 TB volatile pricing data.
Jotai 3.0 Released as ESM-Only Package, Drops Legacy Builds and Deprecated APIs
Jotai, a React state management library, released version 3.0, transitioning to an ESM-only package and removing deprecated APIs and legacy build formats. This update aims to create a leaner core and simplify maintenance by eliminating support for older environments and build systems.
Meta Open-Sources Astryx React Design System with AI Agent Support
Meta has released Astryx, an open-source React design system developed internally over eight years, now in beta. Astryx offers over 150 accessible UI components, customizable CSS design tokens, and tooling for engineers and AI agents, built on React 19 and StyleX. This release provides developers with a comprehensive UI toolkit and introduces native support for AI workflows, potentially influencing how UIs are built and maintained.
Open-Source vphone-cli Enables Full iOS 27 Virtualization on Apple Silicon
The open-source project vphone-cli allows a complete iOS 27 system to run as a virtual machine on Apple Silicon, utilizing Apple's Virtualization.framework. This development provides new capabilities for security research, reverse engineering, and automated iOS testing that were not possible with Apple's official simulator.
Rustls Celebrates 10th Anniversary, Details Evolution, Benchmarks, and Future Plans
Rustls, the Rust-based TLS library, marked its tenth anniversary by releasing a retrospective on its development, community growth, and future roadmap. Benchmarks show Rustls outperforming OpenSSL and BoringSSL in full and resumed handshakes, and in data received throughput, while being marginally slower in data sent throughput compared to BoringSSL.
Netflix Reworks Conductor Workflow Engine for Increased Scale and Efficiency
Netflix re-engineered its Conductor workflow orchestration engine to handle 420 million monthly workflow executions and support workflows 10 times larger. The redesign addresses bottlenecks in workflow evaluation and task coordination, improving performance and scalability for Netflix's internal operations.
tsgolint v7 Stable Release Brings Go-Powered Type-Aware Linting to Oxlint
tsgolint, the type-aware linting engine integrated with Oxlint, has reached its stable v7 release, providing Go-powered TypeScript linting that is significantly faster than ESLint with typescript-eslint. This release offers a performant alternative for developers needing semantic analysis for TypeScript code, covering most type-aware rules and integrating with the official TypeScript compiler port.
Terraform AWS Provider v6.62.0 Adds Support for DSQL, ECS, ECR, SES, Pinpoint
The Terraform AWS Provider released version 6.62.0 in August 2026, adding support for new AWS services including Amazon DSQL, ECS, ECR, SES, and Pinpoint, alongside enhancements to existing services. This update reflects the increasing complexity of AWS infrastructure and Terraform's role as an abstraction layer for managing these services through infrastructure-as-code.
TinyGo 0.42 Adds Recoverable Panics, UEFI Support; Releases Starter Kit
TinyGo released version 0.42, introducing recoverable runtime panics, UEFI target support, and compatibility with Go 1.27 and LLVM 22. These updates close a feature gap with standard Go and expand TinyGo's application to pre-boot environments, alongside a new hardware starter kit.
IETF Introduces New HTTP QUERY Method for Safe Requests with Body
The Internet Engineering Task Force (IETF) published RFC 10008, introducing the new HTTP QUERY method, the first new standard HTTP verb since 2010. This method allows developers to send rich request bodies while maintaining the safe, idempotent, and cacheable semantics of a read operation. The QUERY method addresses limitations of GET requests (URL length limits, data leakage) and POST requests (lack of safety and idempotency) for complex queries.
CPython Adds Official RISC-V Support as a Tier 3 Platform
The Python core development team has officially added RISC-V as a Tier 3 supported platform for CPython, following community efforts to stabilize and validate its implementation. This inclusion in PEP 11 signifies a step towards broader Python ecosystem compatibility with the RISC-V architecture, enabling developers to run Python on RISC-V hardware.
Neovim Adds Native Structured Concurrency Library 'vim.async'
Neovim introduced 'vim.async', a native structured concurrency library within its Lua standard library, to standardize asynchronous workflows. This addition provides a consistent approach for plugin authors to manage tasks, improving stability and preventing blocking of the editor's primary event loop.
HashiCorp Packer 1.16 Adds Native SLSA Provenance Generation and Verification for Machine Images
HashiCorp released Packer v1.16.0, which now natively supports generating, signing, and verifying SLSA provenance attestations for machine images. This update provides a secure, tamper-proof record of how an image was built, addressing a gap in supply-chain security for infrastructure. It matters because it enhances the integrity and traceability of machine images, which are foundational for many workloads, without requiring additional tools.
Netflix Adopts Open-Source Apache Flink Autoscaler for 30,000+ Streaming Jobs
Netflix is migrating over 30,000 streaming jobs to the open-source Apache Flink Autoscaler, moving away from its previous cluster-level autoscaling solution. This change addresses limitations in handling complex, stateful pipelines with varied operator processing requirements, leading to significant cost reductions.
Azure API Management Standard v2 Tier Gains Zone Redundancy
Microsoft has added zone redundancy to the Standard v2 tier of Azure API Management, a feature previously exclusive to Premium tiers. This change allows users to protect against datacenter-level failures at a lower cost, though it requires creating a new instance for existing deployments.
CERN Transitions Accelerator Controls from RHEL to Debian Due to Microarchitecture Mandates
CERN is migrating its particle accelerator control systems from Red Hat-derived Linux distributions to Debian. This change is driven by Red Hat's increasing microarchitecture baseline requirements, which threaten to obsolete CERN's long-lifecycle, specialized hardware.
Figma uses AI agents to automate security investigations and vulnerability detection
Figma's engineering team developed AI agents to assist its security operations, automating alert investigations and vulnerability detection. This system reduced resolution times for complex alerts by 70% and identified over 100 previously unknown vulnerabilities, including critical flaws.
Google Introduces Beyond Zero Security Model for AI Agents, Successor to BeyondCorp
Google published Beyond Zero, a new security model extending Zero Trust to autonomous AI agents, moving access decisions from application to individual resources and actions. This model combines static authorization with dynamic AI-driven decisions to enable machine-speed enforcement for humans and AI agents. It addresses the limitations of BeyondCorp, which assumed human accessors and human-speed actions, by adapting to the scale and speed of AI in enterprise security.
JupyterGIS 0.16 Introduces Collaborative Workflows and Remote Processing for Geospatial Data
JupyterGIS 0.16 adds real-time collaborative editing for Story Maps and vector layers, allowing multiple users to work simultaneously on geospatial projects. The update also integrates openEO for remote sensing processing and jupyter-tiler for large Xarray dataset visualization, enabling interactive exploration of large datasets within Jupyter notebooks.
Airbnb Reduces Authentication Code by 60% with Server-Driven Architecture
Airbnb re-engineered its authentication system to centralize login decisions on a server-side policy engine, reducing client-side code and improving user experience. This change allows Airbnb to adapt authentication strategies dynamically across different platforms and regions without client updates.
Kubernetes Promotes KYAML for Safer, Consistent Manifest Management
Kubernetes is encouraging developers to use KYAML, a stricter YAML dialect, to make configuration more explicit and reduce common errors. This initiative provides a more consistent way to manage increasingly complex Kubernetes configurations without requiring a new parsing ecosystem.
Tether Open-Source Project Brings iOS Continuity Features to Linux Desktops
Developer Zack Bartel released Tether, an open-source project that replicates Apple Continuity features, allowing iOS devices to integrate with Linux workstations. This enables iMessage/SMS, clipboard sync, file transfers, and notifications between iOS and Linux, addressing a long-standing integration gap for users.
Shopify Introduces Gisting for LLM Prompt Compression, Reducing Latency and Cost
Shopify engineering developed Gisting, a technique that compresses large language model (LLM) system prompts into learned "gist" tokens. This method reduces end-to-end latency, lowers infrastructure costs, and increases token throughput without altering core model weights.