From Kotlin milestones to Java AI adoption and Spring AI guardrails
🎯 Featured Article
🎉 Kotlin Turns 15: Celebrate the Kotlin Effect Fifteen years after its introduction, Kotlin continues to shape modern JVM development through multiplatform support, tooling and a vibrant community.
- Celebrates 15 years of Kotlin ecosystem growth
- Highlights community contributions and language evolution
⏩ TL;DR (Quick Recap)
- Java and Kotlin continue expanding the modern JVM ecosystem
- Spring AI focuses on practical LLM safety and production readiness
- AI integration increasingly favors enterprise application platforms
- Developer tooling, maintainability and open source remain key themes
☕ JVM Corner
🚀 Java 27: What’s New? Java 27 delivers a modest release with nine JEPs, emphasizing incremental platform improvements rather than major language changes.
- Reviews every finalized JEP in Java 27
- Explains practical impact for application developers
💡 What’s New in IntelliJ IDEA 2025.2 IntelliJ IDEA 2025.2 improves productivity with enhanced language support, smarter tooling and updated framework integrations.
- Expands support for modern technologies
- Improves everyday developer workflows
🤖 Can You Use Java for AI? Why Java Is Better Positioned Than You Think The JVM is becoming a practical platform for AI integration, especially for enterprise systems where application architecture matters more than model training.
🛠️ How Convenient JPA Defaults Broke Our Kotlin Microservice A real-world Kotlin microservice demonstrates how JPA defaults can introduce subtle performance and correctness issues.
🍃 Spring Updates
🛡️ Spring AI Recipe: Safeguarding LLM Output Build a custom Spring AI advisor that inspects and filters LLM responses before they reach downstream applications.
- Filters sensitive model responses
- Demonstrates custom advisor implementation
🔒 Spring AI Recipe: Applying Semantic Guardrails Semantic guardrails use LLM reasoning to detect prohibited topics beyond simple keyword matching.
- Detects indirect unsafe content
- Strengthens AI application safety
⚠️ Spring Boot End of Life: Every 3.x Branch Is Now Unsupported All Spring Boot 3.x releases have reached OSS end of life, making upgrade planning increasingly important.
- Explains support timelines
- Reviews available upgrade paths
🔍 Extra Reads
Developer Productivity
🧠 Small Programming Tricks — Practical engineering knowledge often delivers outsized productivity gains.
✨ Good Tools Are Invisible — Great developer tools reduce friction instead of becoming part of the work.
📝 Write code like a human will maintain it — AI-generated code still benefits from maintainable software design principles.
⚙️ How I use HTMX with Go — HTMX enables rich interactions while preserving server-rendered Go applications.
AI & Large Language Models
🦙 OpenCode with Llama.cpp — How it Works in Practice — A practical walkthrough of running OpenCode with Llama.cpp under Docker.
🌌 Kimi K3: Open Frontier Intelligence — Kimi introduces a 3T-class open model with long-context reasoning capabilities.
💬 OpenAI Starts Cleaning Up the Utter Mess It Made of ChatGPT — A look at recent ChatGPT desktop improvements and remaining usability issues.
✍️ Stop AI-Shaming Our Precious, Kindly Em Dashes — Please — Challenges the growing assumption that punctuation reveals AI-generated writing.
Open Source & Platform Evolution
🦀 Together for a healthier Clippy — The Rust team shares plans to improve Clippy’s long-term project health.
⚡ Rewriting Bun in Rust — Bun explains the motivations and lessons behind its Rust rewrite.
📖 Microsoft Comic Chat is now open source — Microsoft’s classic IRC comic chat client has officially been released as open source.
Originally posted on marconak-matej.medium.com.