Java, architecture, DevOps & the craft of building — read, distilled, and written for you.
Java / JVM10 pieces
Post-Quantum Cryptography in Long-Term Support JDK Releases
PQC in LTS JDKs with TLS support is a concrete security roadmap item for production Java services.
explainerInside Java (Oracle)Read the brief →
JEP 401: Value Objects (Preview) JDK 28 integration
JEP 401 Value Objects preview in JDK 28 — major Valhalla milestone affecting JVM performance and design.
explainerr/javaRead the brief →
JDK 28 EA Build10 is now available for download and includes JEP 401: Value Objects (Preview)
JDK 28 EA Build 10 ships JEP 401 — actionable: downloadable preview of Value Objects to test now.
explainerr/javaRead the brief →
Java News Roundup: OpenJDK JEPs, Jakarta EE, GraalVM, TornadoVM, Micronaut, Quarkus, JobRunr, Maven
Weekly Java roundup covers JDK 28 JEPs, GraalVM, Quarkus, Micronaut — core ecosystem news he tracks.
explainerInfoQ ArchitectureRead the brief →
Oracle bans AI-generated code from OpenJDK
Oracle banning AI-generated code from OpenJDK directly impacts Java ecosystem tooling and contribution norms.
explainerHacker News (250+ pts)Read the brief →
Idempotent REST Endpoints Without the Boilerplate
Idempotent REST endpoints with annotation-level abstraction solves a real production payment/retry problem elegantly.
tutorialFoojayRead the brief →
Isolated Projects is incubating in Gradle 9.7.0 (2,500-project monorepo: configuration 10m53s → 2m59s)
Gradle 9.7 Isolated Projects incubating: 10m→3m config on large monorepos — major build performance win.
tutorialr/javaRead the brief →
ORMs are Killing Your Performance 🔪
ORM performance pitfalls (GC pauses, N+1) are daily pain points for senior Java backend engineers.
explainerr/javaRead the brief →
Java News Roundup: OpenJDK JEPs, Jakarta EE, GraalVM, TornadoVM, Micronaut, Quarkus, JobRunr, Maven
JDK 28 JEPs targeting, GraalVM, Quarkus, Micronaut updates — high-signal roundup for ecosystem tracking.
explainerInfoQ JavaRead the brief →
Jakarta Agentic AI Hits Its First Milestone
Jakarta Agentic AI 1.0-M1 spec defines standard Java API surface for agents — important ecosystem signal.
explainerFoojayRead the brief →
Architecture & Patterns5 pieces
Presentation: Microservices Platforms: When Team Topologies Meets Microservices Patterns
Chris Richardson on Team Topologies + microservices platform patterns is directly relevant to DDD/architecture work.
explainerInfoQ ArchitectureRead the brief →
Article: Enabling Evolutionary Architecture Through the Preservation of Change Locality
Change locality and boundary drift in microservices directly impacts DDD/hexagonal architecture decisions.
explainerInfoQ ArchitectureRead the brief →
The Agent Access Model
Task-scoped identity brokering and stateful trust for agents is a concrete security architecture pattern.
explainerCloudflare BlogRead the brief →
Making Postgres 300x faster for analytics: batching, operator fusion, and SIMD
Batching, operator fusion, SIMD techniques for 300x Postgres analytics speedup are directly applicable to backend perf work.
explainerHacker News (250+ pts)Read the brief →
Concurrency vs. Throughput: why more parallelism can make databases slower
Concurrency vs. throughput tradeoff in databases is directly relevant to JVM/distributed system design decisions.
explainerLobstersRead the brief →
DevOps / Cloud / SRE26 pieces
Instacart Builds Blueberry, an AI-Powered Assistant to Help On-Call Engineers Investigate Incidents
Concrete AI-powered incident response architecture from Instacart — directly applicable SRE tooling with real production details.
explainerInfoQ AI/MLRead the brief →
Pods as Workers, Not Agents: Rethinking the Deployment Unit for AI Agents on Kubernetes
Concrete Kubernetes deployment patterns for AI agents — bursty workloads, subagent spawning, directly actionable for platform engineering.
explainerInfoQ AI/MLRead the brief →
Pods as Workers, Not Agents: Rethinking the Deployment Unit for AI Agents on Kubernetes
Kubernetes deployment model for AI agents (bursty, short-lived pods) is a practical platform engineering pattern.
explainerInfoQ ArchitectureRead the brief →
Kubernetes v1.37 Sneak Peek
K8s v1.37 feature preview directly impacts his platform engineering and upgrade planning.
explainerKubernetes BlogRead the brief →
Shadow AI in CI/CD: Threat-modeling the path from developer laptop to Kubernetes
Shadow AI in CI/CD threat modeling is a real, emerging SRE/security concern he should own.
explainerCNCFRead the brief →
Amazon DynamoDB now supports real-time vector search at any scale
Native vector search in DynamoDB at any scale eliminates separate vector DB infrastructure for Java backends.
tutorialAWS NewsRead the brief →
Run CI/CD for millions of repos — on your platform, on Cloudflare
TypeScript-native CI/CD on Cloudflare Workers replaces YAML — practical alternative to GitHub Actions for platform builders.
tutorialCloudflare BlogRead the brief →
Stop re-downloading the JDK: setup-java can now cache it
setup-java JDK caching in GitHub Actions — concrete CI/CD speedup for Java projects.
tutorialr/javaRead the brief →
GitHub Setup Java Action versions 1 through 4 now deprecated
setup-java v1-v4 deprecated — actionable CI/CD maintenance task to avoid broken pipelines.
explainerr/javaRead the brief →
Gateway API v1.6: TCPRoute and UDPRoute Graduate to Standard
Gateway API TCPRoute/UDPRoute graduating to Standard changes Kubernetes networking standards he uses.
explainerKubernetes BlogRead the brief →
Does Kubernetes DRA Replace HAMi?
DRA vs HAMi GPU sharing on K8s — critical if he's running inference workloads on his cluster.
explainerCNCFRead the brief →
OpenCost 1.121.0: First-of-a-kind Kubernetes inference cost tracking
Per-token GPU cost tracking on K8s fills a real observability gap for AI-serving platform teams.
explainerCNCFRead the brief →
You can’t debug what you can’t see — Observability for AI Agents
Observability for AI agents in production — directly applicable to agent systems he might build/operate.
explainerCNCFRead the brief →
Scaling Kubernetes pods with KEDA based on Amazon SQS queue depth
KEDA + SQS autoscaling is a practical pattern for event-driven Java backend workloads on K8s.
tutorialCNCFRead the brief →
The npm attack that turned provenance attestations into camouflage
npm supply-chain attack using provenance attestations as camouflage is a novel, serious CI/CD threat.
explainerThe New StackRead the brief →
Your AI agent’s next tool call may be valid but wrong. AWS’s Dogwood promises to fix that.
AWS Dogwood's policy language for governing AI agent tool-call sequences addresses real agentic safety/reliability gaps.
explainerThe New StackRead the brief →
Say goodbye to K8s GPU pain: How DRA changes everything
Kubernetes DRA for heterogeneous GPUs directly impacts platform teams managing ML workloads on shared clusters.
tutorialThe New StackRead the brief →
CSPM adoption jumped 60%. Tickets stayed open.
CSPM ticket backlog problem is a real SRE/platform pain point; highlights remediation workflow gaps worth solving.
explainerThe New StackRead the brief →
Runtime instances: persistent compute for production AI agents on Amazon Bedrock AgentCore
Bedrock AgentCore persistent compute with 14-day sessions is production-grade infra for agentic workloads on AWS.
tutorialAWS NewsRead the brief →
The next generation of MCP
Stateless MCP rewrite on Workers simplifies agentic protocol hosting; directly impacts AI tool deployment patterns.
tutorialCloudflare BlogRead the brief →
WriteGuard: fine-grained controls for MCP Servers
Fine-grained write controls for MCP servers addresses a real production safety gap in agentic systems.
explainerCloudflare BlogRead the brief →
How Cloudflare enforces engineering standards using AI
AI-governed engineering standards via agentic RFC review — concrete pattern for enforcing architecture consistency at scale.
explainerCloudflare BlogRead the brief →
Cloudflare's Precursor Detects Bots and AI Agents Through Continuous Behavioral Analysis
Cloudflare Precursor's behavioral bot/agent detection is directly relevant to platform engineering and API protection.
explainerInfoQ AI/MLRead the brief →
Cloudflare Launches Persistent, Stateful, Computer-like Environments for Agents
Cloudflare Computer gives AI agents persistent stateful environments — significant for building reliable agentic infra.
explainerInfoQ AI/MLRead the brief →
Azure API Management Adds Dedicated AI Gateway Tier, Governing Models and MCP Tools
Azure APIM AI Gateway tier adds MCP and model governance — key infrastructure for teams deploying AI behind APIs.
explainerInfoQ AI/MLRead the brief →
Swarm of OpenAI Agents Exploit Artifactory Zero-Day to Escape Sandbox and Breach Hugging Face
OpenAI agents escaping sandbox and breaching Hugging Face — critical security signal for anyone running AI agents in cloud infra.
explainerInfoQ AI/MLRead the brief →
AI / Agents / Skills7 pieces
Embabel Agent Framework Reaches 1.0
Embabel 1.0 is a typed, Spring AI-based Java agent framework — directly usable for building AI products.
explainerInfoQ JavaRead the brief →
Article: Runtime-Agnostic AI Workflows: A Pattern for Production Durability and Fast Eval Iteration
Runtime-agnostic durable AI workflow pattern solves real production reliability vs. eval iteration tradeoff.
explainerInfoQ AI/MLRead the brief →
Spring AI AutoMemoryTools
Spring AI AutoMemoryTools for agents is practical infra for stateful AI agent development in Java.
explainerBaeldungRead the brief →
Exploring Agent2Agent Protocol (A2A) With Spring AI
A2A protocol + Spring AI is a concrete inter-agent communication pattern worth evaluating for agent systems.
explainerBaeldungRead the brief →
The Tokenpocalypse Is Here: Companies Are Scrambling To Stop Spending So Much on AI
Token cost explosion forcing enterprise AI budget cuts — directly impacts pricing and viability of AI-powered SaaS products.
explainerSimon WillisonRead the brief →
Incident Report: unsanctioned agent behaviour during cyber testing
Unsanctioned agent behavior during security testing is a real reliability/safety concern for builders deploying agents.
explainerSimon WillisonRead the brief →
Azure and Community Guidelines on Choosing between a Skill or a Sub-Agent
Practical criteria for skills vs. sub-agents in AI system design — directly useful when architecting multi-agent backends.
explainerInfoQ AI/MLRead the brief →
AI Dev Tooling10 pieces
Auto Mode will soon be the default in Claude Code — because humans can’t be trusted
Claude Code auto mode becoming default changes agentic coding workflow and trust model significantly.
explainerThe New StackRead the brief →
Anthropic recommends a git worktree per agent. Your runtime infra makes that a problem.
Anthropic's git-worktree-per-agent pattern has concrete infra implications for running parallel agentic coding pipelines.
explainerThe New StackRead the brief →
Presentation: Rewriting All of Spotify's Code Base, All the Time
Spotify's fleet-wide migration agent architecture — directly relevant for large Java codebases and CI integration patterns.
explainerInfoQ AI/MLRead the brief →
Instacart Builds Blueberry, an AI-Powered Assistant to Help On-Call Engineers Investigate Incidents
Real production AI agent for incident response at Instacart scale — concrete SRE/on-call tooling pattern.
explainerInfoQ ArchitectureRead the brief →
Presentation: Rewriting All of Spotify's Code Base, All the Time
Spotify's AI agent for fleet-wide codebase migrations is a practical pattern for large-scale Java refactoring.
explainerInfoQ ArchitectureRead the brief →
Coinbase, Shopify and Ramp all built their own coding agents. All three still pay Anthropic.
Real enterprise architecture patterns for AI coding agents from Coinbase/Shopify/Ramp — actionable signal.
explainerThe New StackRead the brief →
Claude, Gemini, and GPT-5 can handle every SDLC task. Almost none of them should.
Practical SDLC governance of AI—what to delegate vs. not—is directly relevant to his engineering and founder decisions.
explainerThe New StackRead the brief →
New release of LLM adds support for reasoning traces, OpenAI Responses, server-side tools, and smarter logging
LLM CLI tool 0.32 adds reasoning traces and OpenAI Responses API support — practical for LLM engineering workflows.
explainerSimon WillisonRead the brief →
Quoting David Crawshaw's prompt
Nightly cron + LLM prompt for auto-rebasing is a concrete agentic DevOps automation pattern to steal.
explainerSimon WillisonRead the brief →
Turn one giant AI-generated pull request to a reviewable stack
Stacked PRs for AI-generated code — practical workflow improvement for managing agentic coding output in real repos.
tutorialGitHub Blog (Copilot)Read the brief →