Java, architecture, DevOps & the craft of building — read, distilled, and written for you.
Java / JVM14 pieces
Article: Virtual Threads After JDK 24: What Changed for Production Java
Virtual thread pinning fix in JDK 24/25 is production-critical knowledge; downstream saturation patterns replace old bottlenecks.
tutorialInfoQ JavaRead the brief →
JEP 401: Value Objects And JEP 539: Strict Field Initialization Merged Into JDK
JEP 401 Value Objects merged into JDK is a fundamental JVM type system change affecting performance and design patterns.
Breakingr/javaRead the brief →
Implement JEP 401: Value Objects
JEP 401 Value Objects implementation — same as item 9, foundational JVM change for performance and type modeling.
Breakingr/javaRead the brief →
Article: Virtual Threads After JDK 24: What Changed for Production Java
JDK 24 fixed virtual thread pinning on monitors; JDK 25 shifts bottleneck to downstream saturation — critical for his production Java decisions.
tutorialInfoQ ArchitectureRead the brief →
Java News Roundup: Simple JSON API, JEPs for JDK 28, Oracle CPUs, Embabel 1.0, Azul Payara, Helidon
JDK 28 JEPs including Simple JSON API incubator directly impacts his Java backend work.
explainerInfoQ ArchitectureRead the brief →
Newsletter: JDK 27 Rampdown Phase is Closing - Default G1 in All Environments
JDK 27 rampdown and default G1 in all environments directly affects his Java upgrade and tuning decisions.
explainerInside Java (Oracle)Read the brief →
Java News Roundup: Simple JSON API, JEPs for JDK 28, Oracle CPUs, Embabel 1.0, Azul Payara, Helidon
Simple JSON API (JEP 540), JDK 28 JEPs, and macOS/x64 deprecation are near-term roadmap items affecting his projects.
explainerInfoQ JavaRead the brief →
Flamme: A Quarkus extension that treats deployment topology as a configuration, not code. Deploy the same JAR as a monolith or a distributed system.
Flamme lets same Quarkus JAR deploy as monolith or distributed system — directly addresses his microservices architecture work.
explainerr/javaRead the brief →
Transitioning Java to More Frequent Security Updates
More frequent security patches affect JVM upgrade planning and operational security posture.
explainerInside Java (Oracle)Read the brief →
A Bootiful Podcast: Spring Boot founder and lead Phil Webb
Spring Boot 4.1 with gRPC support and MongoDB changes are real architectural decisions for his Spring backends.
explainerSpring BlogRead the brief →
A Guide to Ahead-of-Time Cache in the Java
AOT cache for faster JVM startup directly reduces container cold-start times in Kubernetes deployments.
explainerBaeldungRead the brief →
How to Create a Spring Boot Fraud Scoring Service
Embedding ML scoring natively in Spring Boot avoids Python sidecar — practical architecture win for Java backend teams.
tutorialFoojayRead the brief →
HTTP QUERY Method Explained: RFC 10008, Ecosystem Adoption, and a Quarkus Implementation
HTTP QUERY method RFC 10008 is a real protocol evolution affecting API design; Quarkus implementation is immediately practical.
tutorialFoojayRead the brief →
Ratchet 0.3.1: the CDI-native job scheduler now runs on Quarkus, JVM and native (plus Oracle + SQL Server stores, encryption at rest, cluster coordinators)
CDI-native job scheduler with Quarkus/native support, clustering, encryption — directly applicable to his Java backend work.
tutorialr/javaRead the brief →
Architecture & Patterns10 pieces
Presentation: Clean Architecture for Serverless: Business Logic You Can Take Anywhere
Clean Architecture + serverless vendor lock-in avoidance with Spring directly maps to his hexagonal/DDD and cloud interests.
tutorialInfoQ JavaRead the brief →
How Zalando Built an In-Process Client-Side Load Balancer for One Million Requests per Second
Zalando's 1M RPS client-side load balancer design is directly relevant to distributed systems and high-throughput Java services.
explainerInfoQ ArchitectureRead the brief →
A Detailed Guide to Idempotency, Delivery Semantics, and Deduplication
Idempotency and deduplication are critical distributed systems concerns for any backend engineer.
explainerByteByteGoRead the brief →
Presentation: Parting the Clouds: The Rise of Disaggregated Systems
Disaggregated compute/storage in cloud DBs — directly relevant to his distributed systems and cloud architecture work.
explainerInfoQ ArchitectureRead the brief →
Article: Securing MCP in Production: Defense-in-Depth Beyond the Gateway
Defense-in-depth MCP security layering for production — actionable if he's building or exposing MCP services.
explainerInfoQ ArchitectureRead the brief →
Article: The Hard-Stop Rule: From 3 HCM Monoliths to 120 Domain Microservices
Real 3-monolith to 120-microservice migration story with no dedicated budget — concrete DDD/strangler-fig case study.
explainerInfoQ ArchitectureRead the brief →
Presentation: Clean Architecture for Serverless: Business Logic You Can Take Anywhere
Clean Architecture for serverless with Spring Cloud Functions aligns with his hexagonal/DDD interests.
tutorialInfoQ ArchitectureRead the brief →
Article: An Evolutionary Architecture Pattern for Managing AI’s Pace of Change
AI Gateway pattern for agentic systems is a real architectural concern for enterprise Java backends.
explainerInfoQ ArchitectureRead the brief →
Ontologies Are So Back: Why AI Agents Are Reviving the Semantic Web
Ontologies constraining probabilistic agents within deterministic boundaries directly maps to his DDD/architecture interests for agent design.
explainerLatent SpaceRead the brief →
Article: Securing MCP in Production: Defense-in-Depth Beyond the Gateway
Defense-in-depth MCP security architecture — directly actionable for anyone deploying AI agents in prod.
explainerInfoQ AI/MLRead the brief →
DevOps / Cloud / SRE19 pieces
New Between-Quarters Security Updates for Java: What CSPUs Mean for Your Release Pipeline
New Oracle CSPU out-of-band security patches directly impact Java release pipelines and patching cadence in production.
explainerFoojayRead the brief →
Uber’s Zero Growth Stack: Scaling Services, While Optimising Infrastructure and AI Cost
Uber's infra cost optimization at scale — GC tuning, capacity decoupling — directly applicable to platform engineering.
explainerInfoQ ArchitectureRead the brief →
Kubernetes v1.37 Sneak Peek
K8s v1.37 sneak peek — knowing upcoming deprecations and features matters for platform planning.
explainerKubernetes BlogRead the brief →
Scaling Kubernetes pods with KEDA based on Amazon SQS queue depth
KEDA + SQS autoscaling is a practical, production-grade pattern for event-driven Java workloads on K8s.
tutorialCNCFRead the brief →
Why your company should (try to) build its own AI SRE
AI SRE concept directly intersects his SRE interest and agentic tooling adoption decisions.
explainerThe New StackRead the brief →
Spring Office Hours Podcast: S5E19 - Docker, Compose, Testcontainers, Oh My!
New Docker Compose and Testcontainers features in Spring Boot directly improve his local dev and CI workflows.
explainerSpring BlogRead the brief →
AWS Lambda's Self-Managed Code Storage Lifts the Account Quota, Not the Function Size Limit
Lambda can now use customer-owned S3 for code storage, removing 75GB account quota — unblocks large-scale Lambda deployments.
explainerInfoQ ArchitectureRead the brief →
How the controller-runtime Cache Actually Works, and Why Your Controller Does Not Crash the API Server
Deep dive on controller-runtime cache mechanics is essential for anyone building Kubernetes controllers.
explainerKubernetes BlogRead the brief →
Your Kubernetes health checks are accidentally waking your services. Here’s the fix.
Scale-to-zero health check problem with KubeElasti is a real operational pain point worth knowing.
explainerCNCFRead the brief →
Forget humans “in” the loop. Harness engineering puts humans “on” the loop.
Humans-on-the-loop framing for AI-driven software delivery directly impacts his SRE/platform thinking.
explainerThe New StackRead the brief →
AI-generated software is forcing yet another platform rethink
AI-generated code forcing platform rethink hits his platform engineering and architecture decisions.
explainerThe New StackRead the brief →
Why linting alone can’t govern agentic development
Governance beyond linting for agentic dev is a real architectural problem he'll face shipping AI agents.
explainerThe New StackRead the brief →
“Stateful systems are incredibly hard to build”: How Perplexity thinks about AI agent sandboxes
Perplexity's approach to stateful AI agent sandboxes (Firecracker/microVMs) is directly relevant to platform engineering for agentic workloads.
explainerThe New StackRead the brief →
AWS Weekly Roundup: Local Zone in Athens, Claude Opus 5 on AWS, Lambda durable execution for .NET, and more (July 27, 2026)
Lambda durable execution for .NET and Claude Opus 5 on AWS are concrete AWS platform updates worth tracking for infra decisions.
explainerAWS NewsRead the brief →
AWS Launches Amazon GuardDuty Investigation Agent to Automate Threat Triage
AWS GuardDuty AI agent automating threat triage is directly relevant for cloud security on AWS.
explainerInfoQ AI/MLRead the brief →
Disrupting supply chain attacks on npm and GitHub Actions
Supply chain attack mitigations on npm and GitHub Actions are critical for any production CI/CD pipeline he runs.
explainerGitHub Blog (Copilot)Read the brief →
Tailscale didn't stop the Hugging Face intrusion
Real intrusion post-mortem at Hugging Face exposes zero-trust network limitations — critical SRE/security signal.
explainerHacker News (250+ pts)Read the brief →
The development pipeline is a production system
Treats dev pipeline as production system; strong alignment with his SRE/platform engineering interest.
explainerLobstersRead the brief →
Show HN: Kedge – Full-stack cloud with forkable VM snapshots and global SQLite
Globally distributed stateful serverless with forkable VM snapshots and SQLite — novel infra primitive from a Fly.io veteran.
explainerHN Show HN (100+)Read the brief →
AI / Agents / Skills8 pieces
OpenJDK Interim Policy on Generative AI
OpenJDK's GenAI contribution policy directly affects how AI coding tools can be used in JDK development — relevant to aitools workflow.
explainerr/javaRead the brief →
[AINews] GPT 5.6 price cut by 20%-80%: Cost of GPT 5.4 Intelligence dropped 13x in 4 months due to GPT 5.6 recursive self-optimization
GPT-5.6 recursive self-optimization cutting costs 13x in 4 months via distillation is a concrete, significant pricing/architecture signal.
explainerLatent SpaceRead the brief →
Stateless MCP has recaptured my interest (and inspired mcp-explorer and datasette-mcp)
Stateless MCP 2.0 spec is a meaningful protocol shift affecting how he'd build MCP-based tool integrations.
explainerSimon WillisonRead the brief →
smevals - a small eval suite for evaluating models, prompts, and harnesses
Practical eval framework for models and prompts directly relevant to building reliable AI features.
explainerSimon WillisonRead the brief →
Advancing the price-performance frontier with GPT‑5.6
80% price drop on GPT-5.6 Luna directly reduces cost for AI-powered products he might build or sell.
explainerSimon WillisonRead the brief →
Anatomy of a Frontier Lab Agent Intrusion: A Technical Timeline of the July 2026 Incident
Detailed technical post-mortem of AI agent causing a real cyberattack — critical for AI builders to understand risk.
explainerSimon WillisonRead the brief →
Anatomy of a Frontier Lab Agent Intrusion: A Technical Timeline of the July 2026 Incident
Duplicate of item 2 — same agent intrusion timeline, same high relevance for AI builders.
explainerHugging Face BlogRead the brief →
AI Root Cause Analysis Shifts from Model Reasoning to Context Engineering
Context engineering for RCA pipelines is immediately useful for SRE/observability tooling he might build or use.
explainerInfoQ AI/MLRead the brief →
AI Dev Tooling5 pieces
OpenAI fixed GPT-5.6 Sol’s most frustrating flaw: Burning limits while it waits
GPT-5.6 Sol limit-burning fix directly affects cost/efficiency of agentic coding workflows.
explainerThe New StackRead the brief →
“The beast needs a cage”: Why PortSwigger’s agentic pentesting is kept safe behind bars
Agentic pentesting sandboxing is a concrete security pattern for AI agents he might deploy or sell.
explainerThe New StackRead the brief →
Shipping code without human verification
Automated code verification without human review is a practical challenge for agentic coding pipelines he likely uses.
explainerThe New StackRead the brief →
Stacked sessions and pull requests in the GitHub Copilot app
Stacked PR workflows with GitHub Copilot app are a concrete productivity technique for shipping code faster.
explainerGitHub Blog (Copilot)Read the brief →
The harness is all you need (mostly)
Opinionated Copilot workflow for prototyping to review without tool-chasing is directly applicable to his coding practice.
tutorialGitHub Blog (Copilot)Read the brief →