adam bien's blog

Subscribe to airhacks.fm podcast via: spotify| iTunes| RSS

1. AWS User Group Silesia: Hardcore Serverless Java
   session AWS User Group Silesia Future Processing, Gliwice 24 June 2025
   https://www.meetup.com/airhacks/events/308303637/
2. CloudLand: Hardcore Serviceful + Serverless Java #slideless
   session CloudLand Heide Park Soltau 2 July 2025
   https://meine.doag.org/events/cloudland/2025/agenda/#agendaId.5738
3. LLM / GenAI Patterns, Architectures and Use Cases with Java[online event]
   airhacks.live workshop 10 July 2025
   https://airhacks.live
4. Hardcore Serverless Java on AWS[online event]
   airhacks.live workshop 17 July 2025
   https://airhacks.live
5. airhacks.tv Questions and Answers #livestream[online event]
   live streaming show first monday of the month, 8pm CET
   https://www.meetup.com/airhacks/

Subscribe to airhacks.fm podcast via: spotify| iTunes| RSS

Subscribe to airhacks.fm podcast via: spotify| iTunes| RSS

ZB (Zero Dependency Builder) is a simple build tool for Java without external dependencies, Main branch only development works well for most enterprise projects but not when releases are dictated by law, Load balancers like HAProxy and Nginx are recommended for on-premise while cloud load balancers are good for global distribution, JSON-P is a simple way to handle JSON in Java applications, Entity mapping can be done with direct serialization or through DTOs depending on API versioning needs, EJBs can be faster than RequestScoped beans due to object pooling and graph reuse, Quarkus deploys at build time and performs tree shaking for better performance and security, Native images make sense for on-premise deployments but not necessarily for cloud/serverless, AI tools are useful when you can quickly evaluate and accept/reject suggestions, Trunk-based development avoids merge conflicts and keeps the codebase in a functional state, Virtual threads make reactive programming less necessary

and answered questions:

1. Why should you use static method in an interface why not just use a class?
2. How are you using AI tools like copilot or cursor?
3. What is a good load balancer?
4. How would you map in a project that use BCE an entity post from block?
5. How would you do if you want a different request for the creation for the post as well?
6. Would you create the attribute names with static strings?
7. Are you aware of JSON-P being brought into the JDK?
8. Would you like to update the benchmark about request scoped and EJBs?
9. Do you use native images? Why or why not?

Ask questions during the show via twitter mentioning me: https://twitter.com/AdamBien (@AdamBien),using the hashtag: #airhacks or built-in chat at: airhacks.tv. You can join the Q&A session live each first Monday of month, 8 P.M at airhacks.tv

1. Meet zb (zero dependencies builder)
2. Request for recommendations for load balancer tools for on-premise applications, with previous GSLB experience being limited compared to cloud-based solutions and seeking more feature-rich alternatives.
3. Use DTOs like PostDto in the Boundary layer for responses to avoid exposing entire JPA entities, and utilize separate DTOs like CreatePostDto for handling creation requests.
4. Boundary layer manages interaction between application and external systems, handling requests and responses using appropriate DTOs.
5. Managing CI/CD and collaboration with multiple people using only the "main" branch workflow.
6. Comparing RequestScoped and EJB performance in Quarkus, questioning necessity of waiting for "EJB Replacement" for performance optimization.
7. Discussion on whether RequestScoped is currently sufficient without optimizations like pooling and max concurrency that are only available in EJB.
8. Payara's startup and deployment speed now comparable to Quarkus thanks to Project Leyden and JDK 24 improvements.
9. Java's overall performance enhancement with JDK 24 reduces performance gap between Quarkus and other Java frameworks.
10. Project Leyden makes reflection no longer a performance bottleneck, benefiting all Java applications.
11. Introduction of jtaccuino project - a Jupyter notebook written in JavaFX providing an alternative to Python for datascience.
12. time travel, topics form the 35th episode:

    Java EE 8 News, Java SDK for Oracle Bare Metal Cloud Services, Handling JPA relationships in CRUD microservice scenarios, The role of RxJava in typical enterprise applications, JAX-RS returning Response vs objects and JSON vs XML, Cross-field entity validation, GPL/LGPL opensource and commercial projects, What is pac4j.org, Transactions CRUD-Services BCE EntityManager and connection sharing, SCM Managing Structuring and organizing microservices, Gradle adoption gradle vs maven, Dealing with setup scripts in Java EE environment, Oracle JDK vs openJDK and licensing, Greetings from Panama jmoordb, The role of Model Driven Architecture in Java EE

Any questions left? Ask now: gist.github.com/AdamBien/74fcde9b08475112003c772ed8384462 and get the answers at the next airhacks.tv. Some questions are also answered with a short video: 60 seconds or less with Java

Ask questions during the show via twitter mentioning me: https://twitter.com/AdamBien (@AdamBien),using the hashtag: #airhacks or built-in chat at: airhacks.tv. You can join the Q&A session live each first Monday of month, 8 P.M at airhacks.tv

Subscribe to airhacks.fm podcast via: spotify| iTunes| RSS

Subscribe to airhacks.fm podcast via: spotify| iTunes| RSS

Subscribe to airhacks.fm podcast via: spotify| iTunes| RSS

Subscribe to airhacks.fm podcast via: spotify| iTunes| RSS

"This interactive session is about practical, no-nonsense coding. We'll explore lean Java 21+ practices and patterns that eliminate bloat and hacks that increase productivity. We will start with structuring and organizing backend monoliths and microservices, cover testing, then move on to decoupling, abstractions, Data Oriented Programming, automation with pure Java."

I used the following example during the session ebank. As a quickstarter for serverless / AWS production projects, I'm using the following template: aws-quarkus-lambda-cdk-plain

Subscribe to airhacks.fm podcast via: spotify| iTunes| RSS

REST is about manipulating resources with well-defined HTTP methods (GET, POST, PUT, DELETE) while gRPC is just remote method invocation, Quarkus request-scoped beans are faster than in traditional Java EE due to lack of reflection, Java is 35x faster than Python making it better for AI inference and production, serverless architecture provides cost benefits by paying only for actual usage, RAG (Retrieval Augmented Generation) can be implemented in Quarkus with predefined pipelines, Neo4j works well with LLMs due to its graph structure matching semantic relationships, document categorization with AI can be implemented by passing document chunks to embedding models and using vector databases, Java is getting better with each version with Java 24 being particularly strong, event-driven architecture works well for AI applications by chaining asynchronous operations

See you every first Monday of the month at https://airhacks.tv 8pm CET (UTC+1:00). Show is also announced at: meetup.com/airhacks.

Are you nice? :-) Then checkout: the airhacks discord server

"Treating LLMs as microservices with APIs for enterprise integration, Java is excellent for model inference and production deployment, Enterprise applications benefit from LLMs as decision support tools rather than decision makers, Langchain4j gaining adoption in the Quarkus enterprise space, Java developers and Python developers form distinct groups with different priorities, Future trend toward smaller dedicated models through model distillation, Java's vector API can be as fast as modest GPUs in some cases, Local LLM inference possible with Java at 10 tokens per second, Java 24 released with gatherers and FFM API, Java One conference was well-organized with great content and networking, JavaLand conference focused on community with keynote about 30 years of Java, Centralized logging now standard in cloud environments, Application-scoped services preferred in serverless architectures, Generic utility packages should be avoided for better cohesion, Preference for JSON-P over JSON-B for better versioning capabilities, Synchronous communication preferred over queuing when possible, Vaadin making a comeback with web components support"

See you every first Monday of the month at https://airhacks.tv 8pm CET (UTC+1:00). Show is also announced at: meetup.com/airhacks.

Are you nice? :-) Then checkout: the airhacks discord server

1. A more straightforward explanation of the difference between gRPC and REST regarding query APIs
2. Performance of RequestScoped on Quarkus. What about missing EJB features?
3. What are some interesting applications of AI Agents and automation using Java/Quarkus?
4. How can one effectively integrate AI Agents and automation with Quarkus?
5. What advice do you have for someone new to AI Agents and automation who wants to work with Java/Quarkus?
6. How does Quarkus compare to Low-/No-Code interfaces for AI Agents and automation?
7. What are the benefits of integrating AI into the correspondence exchange system?
8. How will AI be used to read, summarise, and categorise documents in the system?
9. What are the challenges of using a local LLM for document processing?
10. How does the system ensure cost-effectiveness by using a local LLM?
11. Time travel, topics from the 34th episode:

    The state of microprofile.io, Motivation for StatusTest, Java EE -- feedback from JUGs, Micro, Swarms and co. the possible killer use case., SPA and JAX-RS. One or multiple projects?, Files filtering and Java FX, REST architecture in 5 mins, Should we keep SOAP?, Personal opinions on article: "Oracle finally targets Java non-payers", CDI with transactions, When should we start to worry about performance?, How we can communicate multiple microservices without share their DTO's?, A lean way to inject properties with Java EE., Java EE vs. Spring, Is it possible to put HTML5 application after the .war has been compiled?, Future of NetBeans, Non blocking IO vs. JDBC, Java EE and scalability, How to connect SPAs with backend efficiently, Java EE vs. Spring Cloud -- how to deal with non functional requirements, Can we start threads in a @Singleton?

Any questions left? Ask now: gist.github.com/AdamBien/df41e421a58f9f8ef0963733f6019725 and get the answers at the next airhacks.tv. Some questions are also answered with a short video: 60 seconds or less with Java

Ask questions during the show via twitter mentioning me: https://twitter.com/AdamBien (@AdamBien),using the hashtag: #airhacks or built-in chat at: airhacks.tv. You can join the Q&A session live each first Monday of month, 8 P.M at airhacks.tv

Subscribe to airhacks.fm podcast via: spotify| iTunes| RSS

Subscribe to airhacks.fm podcast via: spotify| iTunes| RSS

• Android Studio
• Aptana Studio
• Arachnoid
• Atom
• BBEdit
• BEA WebLogic Workshop
• BlueJ
• Borland JBuilder
• Brief
• Cloud9 IDE
• Codeanywhere
• Code::Blocks
• CodeWarrior
• Codenvy
• CodeLite
• CodeMix
• CodeRunner
• ConTEXT
• Crimson Editor
• DrJava
• Eclipse
• Eclipse Che
• EditPlus
• Emacs
• Ensime
• Geany
• Gedit
• Gitpod
• Greenfoot
• IBM Rational Application Developer
• IntelliJ IDEA
• JCreator
• JDEE (Java Development Environment for Emacs)
• jEdit
• jGRASP
• jMonkeyEngine SDK
• NetBeans
• Notepad++
• Oracle JDeveloper
• Processing
• PSPad
• SciTE
• SciTools Understand
• Spring Tool Suite (STS)
• Sublime Text
• SmartGit
• TextMate
• TextPad
• Together
• UltraEdit
• Vim
• Visual Studio
• Visual Studio Code
• Xcode
• XEmacs
• Zeus
• Microsoft Visual J++ (Historical)
• Symantec Visual Café (Historical)
• IBM VisualAge for Java (Historical)
• Sun Java Studio Creator (Historical)
• Sun Java Workshop (Historical)
• Sun Java Studio Enterprise (Historical)
• Forté for Java (Historical)
• Borland JBuilder (Historical)
• SuperCede (Historical)
• CodeWright (Historical)

Subscribe to airhacks.fm podcast via: spotify| iTunes| RSS

• Apache Tomcat
• Apache TomEE
• Apache Geronimo
• Apache JServ
• Eclipse Jetty
• WildFly (JBoss AS)
• JBoss Enterprise Application Platform
• IBM WebSphere Application Server
• IBM WebSphere Application Server Community Edition
• Oracle WebLogic Server
• GlassFish
• Payara Server
• OW2 JOnAS
• Apache MINA
• Netty
• Borland Enterprise Server
• Adobe ColdFusion
• Lucee (Railo)
• Oracle OC4J
• Orion Application Server
• Caucho Resin
• SAP NetWeaver Application Server
• TmaxSoft JEUS
• Fujitsu Interstage
• Hitachi Cosminexus
• ATG Dynamo Application Server
• SilverStream Application Server
• Sybase EAServer
• iPlanet Application Server
• Sun Java System Application Server (Sun ONE)
• Netscape Application Server
• NetDynamics Application Server
• JBoss Undertow
• Spring Boot
• Dropwizard
• Eclipse Vert.x
• Micronaut
• Oracle Helidon
• KumuluzEE
• Open Liberty
• Thorntail (WildFly Swarm)
• Enhydra Server
• NEC WebOTX Application Server
• Persistence PowerTier
• Secant ModelMethods
• TradeCity Rexip AppServer
• Trifork Application Server
• HHPN XLiRAD
• Interactive Business EAS
• New Atlanta ServletExec
• Gefion LiteWebServer
• Red Hat Quarkus
• Sun Java Web Server
• Jeeves
• GemStone/J (GemStone Application Server)
• Blazix
• IronFlare Orion Server
• Pramati Server
• WebGain Application Server
• BEA Tuxedo (Java-enabled)
• JetSpeed
• WebObjects
• Jo! Application Server
• Dynamo Application Server
• Simple Web Server (SWS)
• Winstone Servlet Container
• nanoHTTPD
• TJWS (Tiny Java Web Server)
• Undertow.io
• Grizzly HTTP Server
• Restlet Server
• Akka HTTP
• Play Framework
• Ninja Framework
• Spark Java
• Ratpack
• Jooby
• Jodd Madvoc
• Vert.x-Web
• Pippo
• Struts2 embedded server
• RestExpress
• JLHTTP
• Simple Framework
• Virgo Server
• Piranha Cloud
• Siwpas (Simple Web Profile Application Server)
• Tengah

The list is also available at: java-archives

1. JavaOne report
2. JavaLand report
3. What stage are people at wrt writing GenAI applications in Java, or integrating AI use cases into their existing Java apps. Are they learning, prototyping, developing, using in production ?
4. Are you seeing an increase in activity in Langchain4J adoption recently to solve the use cases that might have defaulted to being written in Python?
5. Any interesting use cases you can share on GenAI applications in Java being worked on in the wild?
6. 100 episodes back in time:

   Java EE 8 news, Managing logs in central location, About layering: @ApplicationScoped for services?, Why I created an ExceptionFactory--a long time ago., Util packages and classes, Modal windows in JavaFX and background threads, Managing JDBC drivers in Docker, How to run a DB in a container?, To share or not to share a DB, Managing private docker images, DTOs vs JSON-P for partial data transfer, Microservices: queuing choices -- JMS or not, Rolling updates with Java EE and docker, Dealing with bloat: lombok, IDE or what?

   33rd airhacks.tv questions

Any questions left? Ask now: gist.github.com/AdamBien/13fff4f53cf668c9cdfe87ff551bf61a and get the answers at the next airhacks.tv. Some questions are also answered with a short video: 60 seconds or less with Java

Ask questions during the show via twitter mentioning me: https://twitter.com/AdamBien (@AdamBien),using the hashtag: #airhacks or built-in chat at: airhacks.tv. You can join the Q&A session live each first Monday of month, 8 P.M at airhacks.tv

Subscribe to airhacks.fm podcast via: spotify| iTunes| RSS

Subscribe to airhacks.fm podcast via: spotify| iTunes| RSS

1. Convenient and nice location: all sessions were in one building - there was enough time to walk between sessions, have a hallway conversation ...and get a coffee.
2. Speaking of coffee, the standard coffee was good and much better than in the hotel. And great Italian speciality coffees were served by passionate San Francisco baristas. I had my first Cortado at JavaOne.
3. Excellent presentations by core Java/Oracle engineers. They were well prepared with great content. Talking to (former) Oracle engineers / Java committers has always been my main motivation for attending JavaOne.
4. Good (great?) food, (too) good local biscuits. Always plenty of drinks and coffee.
5. Great attention to detail: useful gifts, nice stickers, JavaOne branded seat cushions, Java coasters, JavaOne signposts, badge flags and posters.
6. Good party venues: the food was good, the venues were nice, and it was easy to make conversation.
7. JavaOne was also better attended than expected.
8. Perfect organisation: from the badge to the schedule to the conference app, everything worked perfectly.

I was surprised by the quality of the conference and started asking attendees about their experiences. The answers were all positive.

The most surprising fact was that the majority of people were attending JavaOne for the first time. I asked during my presentation - more than 70% had never been to JavaOne before.

It is hard to find anything that could be improved, other than the size of the rooms. The rooms were either too big or too small. There was a lack of medium sized rooms. I also missed the evening BoF sessions, but my schedule was already packed.

In the event that my proposal is accepted - see you again in the year 2026

Subscribe to airhacks.fm podcast via: spotify| iTunes| RSS