Foojay Podcast #69: All Things Java at VoxxedDays Amsterdam

Transcript

Automatically generated from the audio, so it can contain errors in names and technical terms. Suggest a correction.

[0:00] Let’s talk about all things Java at Fox Days. Welcome to the Foojay podcast. All your news about OpenJDK. On April the 3rd, the first Vox Days event in Amsterdam took place. Vox Days are tech events organized by local community groups with support from the Devox team. Han willing brought along a camera and microphone and spoke with many of the attendees. But let’s start with Kurk, one of the events organizers who was very happy with how the day started. Congratulations. It seems like a big success, a big crowd, a lot of people talking, a lot of enthusiasm here. Yeah, thank you. Really nice to hear and especially the people if they’re having a good time, the talks are good, having a great time, social, I’m very happy with the JS also with Foojay, everything comes together. I’m really I’m proud of that of us, I think. Yeah. I mean this took many months to put together and a lot of people are involved and it must be fantastic. Great weather and so many people you know wonderful. Yeah thank you. Thank you. Yeah the weather is also helping and of course we have a jug square. How cool is that all

[1:10] The jugs are here and yeah it’s all comes together kind of and is there anything that you can predict about future implementations of this interface of this interface. Interesting. Yeah we want to grow also next year. Next year we are a two-day event, Fox Days event on the 25th and the 26th of March. So we have even two days of fun. Yeah. And to grow of course. Vox Days is part of the Devox family which was co-ounded by Stefan Jansen in Belgium many years ago. So he had a very good reason to be in Amsterdam. This is the inaugural event. I have to be here, right? Just to see it in my own eyes how it’s going and so on. It’s just amazing. I mean, congratulations to the team. I mean, they did an amazing job. I mean, I already feel the Devox vibe. I mean, it’s in the cinema complex. They have almost a thousand attendees. it’s even the AV team is the team that also does Devox Belgium. So, it’s clear to say we know clearly what their ambition is. is to become a DevOps Netherlands and they’re getting very close. so the criteria is to have at least minimum a thousand attendees be

[2:19] Two days minimum and so next year Fox Day Amsterdam will be two days which is already great. So the chances are very high that not next year but the year after they will most likely become a devox Netherlands which was the initial goal by Paul Kra and his team. So I can only congratulate them. I mean it’s it’s amazing. I mean it’s the first Fox days which has already 950 people in their first edition. So they are skyrocketing. that’s for sure. It’s it’s really amazing and also the number of speakers and the quality of the speakers and the variation of new speakers and well-known speakers is also quite great. Well, so they I think they had close to 500 proposals in the CFP and you know if you only well only if you have like 30 slots you have a huge number of talks that you can select from. I mean I even submitted the talk and I didn’t get selected. That’s how high the quality was right. So which is great. I mean content is king. That’s really important. and we have like you know Josh Long is here, Victor Renta is here and many other well-known speakers which of course it helps the

[3:25] Branding the franchise really helps because if they see ah there’s a Vox days the speakers also know that we have top quality and good content. and that really helps right if you start from scratch with an unknown brand you would not get that many submissions because people don’t know about it. So it looks like this McDonald franchise thing. It looks like it works. And also it may have something to do with the community in the Netherlands which has always been quite strong in the Java ecosystem. A lot of different jugs and always great attendance at the deboxes around the world and so on. Yeah. Well, I mean Turk really did an amazing job. He went to all the Java user groups in the Netherlands and there’s so many I wasn’t aware that were there so many. So even you guys as well, he really reached out. They gave them even a floor space to promote their Java user group and it’s all about community. I mean this is only possible with the support of the community with support of the sponsors of course because they have a lot of sponsors. I mean they sold out the

[4:23] Exhibition floor and there were so many new companies that we’ve never seen in the DevOps Vox days ecosystem. So there’s a lot of Dutchon companies here as well which is great. so yeah it’s it’s very healthy, great venue, cinema, good speakers, community. It’s it just matches all and you see that by just experiencing this event. It’s it’s really fantastic and congratulations for your part in it and it didn’t do too much. I mean it’s really Paul Krux his team and I mean if you look at the volunteers it’s really congrats to them. I mean it’s it’s wonderful. Thank you. This is the first Foojay podcast ever to feature more than 20 guests here. Tion asked the same two questions to many of the conference visitors. Tell us who you are and what excites you about the technology landscape and what are two tips or insights you’d like to share. As you might expect, there’s a lot of talk about AI and machine learning, but you’ll also hear about new Java features, profiling, open- source security, code reviews, and much more.

[5:26] Let’s get started. I’m Litzk Leo and I’m very interested in Java 24 and AI and can you tell us one or two things that you find especially interesting in that area? I think that the most important ones are the features that doesn’t need any code chains and that you can just apply with a simple Java flag and then already see performance and what which things are those can you tell us some of those things? yes, the Jeep 450 the compact object harness that’s one of them and just a simple flag added to your startup and you already see profits and can you tell us something else you find interesting in that context? yes there are lots of new features with Java 24 and there ahead of next to the 450 there is also a flag for the head of time class loading and linking and that’s now without pinning and that’s also very interesting to do.

[6:25] I’m Johannesburgger I work at SAP and I’m an OpenJK developer so I work on Java itself. I focus mostly on profiling and like on developer tooling like on like at this conference where I’ll be giving a talk on instrumentation. I think in it’s important to know that it’s not a magic that you could wield and that solves all your problems but you should understand what your tools are doing. The second thing is there will be a cool new profiler coming hopefully in JDK25. Let’s see maybe in JK26 that will be a CPU term profiler directly integrated into JFR. I’m working on this now for like three years. So fingers crossed. You are Christian Solv the Spring AI project lead. I am project lead. I am co-lead for the Spring AI project and I’m founder and lead for the MCP Java SDK project as well which is something that just started now. This model context protocol this new fancy new stuff that help us build AI applications with Spring ecosystem. If you want to make generative AI useful your domain generative AI stateless frozen pre-trained sometime in the time some somewhere in the past. So you have to

[7:31] Make the generative AI aware of your data. You have to make the generative AI aware of your API how to call your API. Genifi takes text in textile but actually you can make it work and then you have to take care for the memory state. You have conversations and that’s all this is a integration problem. Spring is pretty good in this spring has this element ecosystem. ecosystem that would make you very very natural using it in building your AI applications. And now we have this new kit in the block called model context protocol. If you see yourself building a lot of tools in order for your AI application to interact with this environment, you can find the model context protocol as the best way to provide unified access to tools that some someone else built in some other platforms and use it within your spring environment or vice versa. you can actually expose your own tools and let any Python framework or JavaScript framework that use AI use it and it’s pretty nice time to be Yes.

[8:26] And actually also I saw you do a demo integrating cloud AI with some Java code. How did you do that? Yeah, this is a very good question. This is follow up of using model context protocol allow us to unify and connect a client in this case cloud desktop which is a MCP client to talk to a server. In our case, it was a server built in Spring Boot application, Spring Boot application. And you can actually have your code desktop talk to this server and get some information in real time. Actually, you allowing the cloud model through the client to interact with your Java application in your that you build your Spring Boot. I’m Tom Kohl’s. I’m one of the by now co-organizers of the Belgian Java user group because I got some friends that now help out. and my main interest in the space right now is anything involving AI. And when I say AI, I don’t mean the narrow focus of Genaii, but actually everything that’s more classical AI like machine learning, metauristics, mathematical optimization, all these technologies that have existed like since the 70s and 80s that are technically also AI, but a

[9:33] Lot of people don’t consider them AI anymore because they now understand how it works. And what I noticed is that the definition of AI seems to shrink every time we understand it. When it first when we first looked at Pong like the video game, some people considered it, wow, it can react to how I’m playing and I’m playing against AI. Then we said, yeah, no, that’s too simple. That’s not AI. then we made complexer algorithms. But once we understood those, people said, yeah, no, that’s not AI because it doesn’t learn anything. Then we made machine learning. We gave it a bunch of data and it started learning. And then we said, yeah, but even machine learning, that’s not really intelligent. It’s not creating anything. Now we’re in the space of Genai. It can create stuff, but now people are saying, “Yeah, but it’s not intelligent.” So the term AI, it seems like we as as a human species, we don’t want it because every time we have something that looks artificial intelligent, we say we kind of change our definition of what does intelligent actually mean. So it keeps moving and moving. So I’m very interested

[10:29] In Genai because I think it’s the most fascinating tech that’s right out there right now, but to integrate it with all this other stuff around it that has proven itself for many many years. So that’s where my focus is right now. I would like people to think a bit more that when someone utters the words I use AI to really question what form of AI is being used here because often they mean Gen AI especially when people say I’m coding with AI. They often mean it’s geni. It’s LLM generating but there’s so much more to the AI space and I feel that we’re losing some of that very valuable technology that’s in the other layers of the old school AI. so that’s the first thing and the second thing is once you make that move and you investigate that you will see that there’s so much opportunity to integrate these these older forms of AI with the new genai hype and if you integrate those you can have some really fascinating really powerful solutions for problems that are everywhere. I’m Eric Bulamea I’m a developer at moment in Java and CGI.

[11:34] What area of technology of an interest at the moment is automated code analysis. So basically going through huge code base and seeing where the problems are, how we can improve it because projects are getting bigger and it’s harder and harder to add features or remove bugs. What are one or two things that you find especially interesting or relevant about this to you? Well, that there actually are ways for example to use git and github analysis to see what pieces of code change frequently and which change in tandem. So you know these two things are dependent or freedom things are dependent and they probably shouldn’t be. So it gives information about where should refactor for example. So if you have lot of technical depths which most companies have where to start with technical depths are you just keeping things up that are unimportant or are you going through the to the regulum to the heart and is this something that GitHub provides or where does this come from? basically you can do analysis with git. So if you look at books like your code is a crime scene by

[12:43] Them tornill he has actually made tools but you can also program them yourself if you not need to. So you can make analysis which codes change files change together and which files change frequently and together with a complexity analysis how big the file is all the measures you can say we should address that first. My name is Abraham founder Feifer I currently what I find interesting interesting in the industry is of course the GI movement and I’m curious where this is everything is heading at the moment and what two things in particular concern you or have you been thinking about in that area? Well, the first thing of course is how it will impact my job as a software developer and secondly I’m actually curious of what impact it will have on the open source community or well open source projects in general. So what impact do you think it could have?

[13:30] What’s the connection between open source projects and AI? Yeah, I think we’re pretty much at this moment we’re thinking a lot about reusing code like somebody else wrote a library and you’re implementing it and a lot of times with software we see that only a small slither of that is being used and now with like to extent like I’m wondering if this vibe coding movement as well will force developers actually to write that specific slice of that library themselves. and yet what impact that will have in general on the open source community like if there’s certain libraries that will actually stop existing in a way just because people will be writing it themselves like I think in the nojs communities anyway we’ve seen like where people would like use simple like string concatenation and they will include all of these tons of libraries but now actually they can just ask their junior developer to write that library for them in this case an LLM or a code assistant.

[14:23] So does it mean the end of junior developers? I don’t think it’s the end of junior developers but I do think the way that we are skilling up people will be challenged because for example if you now can if you would take time with a junior developer to explain to them how to write code now you can also do that with a smart agent and if the need for those junior people gets less the question is how do we get people into the industry though of course the junior developers will also be well good at writing code but I’m not sure if they will be good at evaluating the quality of that code which is yeah we’re here to explore that well the future will tell us about that my name is Shham I work as a micro in Microsoft as a pre-sales consultant or pre-sales technical architect and my focus area is combining cloudnative applications with AI large language model or small language models I find the interesting area is obviously and of course the genai side the way it is evolving the innovation pace of innovation. It’s mind-blowing. So that’s my core and that’s why I’m looking at in

[15:31] Any conference or any talk I’m actually going. I still think u quite a lot of us read about this listen go to talk to about this but we don’t really use it as much. the first thing I’m going to talk about is if you have a GitHub account I think everybody does at this point of time GitHub gives you a way to use these models freely. So if you go to GitHub marketplace, you can search for the models that you want to work with. you can in their playground do prompt and see how this model performs. You can compare these models. But on top of that you can get a developer key which is completely free has a usage limit but still completely free and you can use that development key to write code against it. So use langen 4j, use semantic kernel, use your favorite framework or even no framework as well. Spring AI as well write code on it, see how it works and get familiar with it. I mean to understand a technology you need to use it you need to know how to do it. So that’s the one thing I want to give. Second thing I want to give is be critical about technology. Every new aspect or a new thing is very shiny

[16:41] And people mostly talk about the shiny side of it. but be critical what it cannot do, what it can not bring it to you and how you will going to tackle that. So be crit critical about the advancement of the technology and get to know it properly and then yeah you would be I think the best person to judge whether this is good for you or not. That’s it. My name is Josh Long and I work on the Spring team as a developer advocate and what I’m I think I’m not the only one surely I find AI super interesting right I mean it’s it’s sort of this very pervasive thing right now and I think it’s just been monopolizing a lot of my time you know I mean obviously there’s first of all pierce through the hyperbole a little bit if you can that’s hard to do but if you can do it it’s and you can find the things that are actually really truly valuable that is very interesting Right.

[17:35] And I think right now we’re seeing a lot of people hit the edges of some of the potential that they were sold and they’re saying, “Okay, this isn’t real, but if you go back a few steps, there’s a real thing there.” So, one thing I’ve always wanted was the Star Trek computer, you know, computer, you know, simulate this or blah blah blah, you know, and now you can kind of have that, right? You can actually talk to your computer and get interesting things to happen. And for example, there’s this new MCP protocol which by which you can drive whatever you want other services, APIs, local software. I mean like in theory, it wouldn’t be hard to build an app that says computer render a scene of a cafe in Blender and then it just it’ll do it, you know? That’s crazy. That’s that’s orders of mag. It’s a your voice is your for all these years we always spent time trying to find these four jails you know these things that the business analysts could use cuz in that narrative in that reckoning apparently business analysts are idiots that’s that’s the way it’s always seemed to me

[18:37] People are like oh SQL so you don’t have to actually learn to write code right well why anybody can write code you know it’s not hard but anyway that said I think the fourth gener the 4GL that we all all been looking for is English or Dutch or whatever right like that’s the language and it’s it’s here you Human language is a new programming language. Somebody said to me recently, “Oh, well that’ll that person sounds smart.” Yeah, that Yeah, that person was Christian Sulliv. Oh, that guy is the best. Yeah. Yeah. Yeah. Yeah. Super powerful. So, do you really see this happening that the human language will completely operate the machine? No, but I do see that you can do so yes, I see it’s going to happen. Yes, I don’t see it happening right now, but that’s only because the protocol that I think will enable it is very very new. It’s MCP, right? and it’s basically a way to interface your model with your disparit data services systems and so on and those things are it’s it’s sort of all the things we wanted from SOA 25 years ago without WSAR without the concerns around schema around

[19:38] Integration orchestrate all that just goes away because you got the model doing it and it understands English it can inspect the shape of your function of your service and deduce how to call it just the schema is the language description of Actually the schema could be the Java do you know like something like that you know what is the role do you think of the junior developer in the future if if AI is going to be replacing all these grunt kind of tasks and increasingly more and more of these where does the especially the junior developer remain in the story how do you actually learn programming skills or are you saying that you will not need programming skills no I think it’s up I think these MCP integrations just give you a higher order sort of abstraction, right? A higher order bit that you can work with.

[20:25] So, you know, in the same way that object orientation didn’t didn’t stop programming quite the contrary, we gave it a force multiplier, the level of abstraction moves up, the more a competent person can do, you know, the force multiplier of that is more and more, you know, as we get better and better technology. So, yeah, I think the younger engineer today has a unique opportunity like learn the basics. Obviously, you should always know what you’re dealing with before you start abstracting away. But then, my goodness, I mean, to be able to sit at top that perch with a fresh fresh perspective and to be able to build. I mean, do you remember when you were young and you had that like, well, I can build anything and nobody has told me otherwise, so I’ll try it. And you sometimes you sit sit down, it’s like you sit down for a month and you come out the other side of it and go, “Oh, that was insane. I should never have done that, but I’m glad I did.” Right?

[21:14] That’s because you didn’t know how hard it was going to be and you just dove into it and the next thing you know, something cool happened. Right? Imagine having that perspective, that lack of inhibition, but being able to wield AI. I mean, the world’s our oyster right now. You’re if you’re a young person, you have that this toolbox at your disposal. It’s a great time. I’m Sona Peterson and yeah, I’m I’m a Java developer but also very busy with Rack and AI. So, yeah, that’s that’s what I’m interested in at the moment. And you’re also one of the community reviewers for this conference Vox Days Amsterdam. Can you tell us about that? What that means? Yeah. when Fox Days Amsterdam was announced, they said that you could become a community reviewer, which meant that every proposal for a presentation would be available for you to review and you could do like a pre-selection and advise the organizers which one to select.

[22:20] And that was a lot of fun. There was so much and well they didn’t listen to me of course in everything otherwise it would been a one rag u conference and that was not the idea of course but some of them come came true and that’s why I’m here. It’s fantastic. And about rag are you doing anything with it at the moment and what are your main insights or surprising things that you’ve discovered? Yeah, I’m working on a vector search application and we also experimented with rag in a chatbot and today I went to the presentation from Gom Lavfor about how to use rag better and well I went there in to his presentation in defocent rack and well I’m I’m really looking forward to try it out tonight.

[23:22] My name is Brian Vir. I am the co-lead of the Netherlands Java user group. I work for Sneak the security company and currently what we’re working on is obviously something with generative AI and LLMs but the security angle on using LLMs and what could possibly go wrong and that is a super interesting part that I’m currently working on. So what are two things from that you’ve learned recently or that you find inspiring or interesting or that you would leave as tips about that topic? Oh, there are a bunch of things. First of all, it’s hard to secure it. I mean, prompt injection is there are so many forms of it. but the fact that LLMs now can have tools or even aentic agentic behavior, that opens up a whole another ball game from a security angle.

[24:07] What could possibly go wrong? So things are what you could prevent is hey check your in and output what goes into an LLM maybe for instance with guardrails currently that’s for instance implemented in LangChain4j with Quarkus but there are other things as well like regular vulnerabilities like a SQL injection or a path reversal problem can propagate into things that are getting injected into an LLM while and that makes an LLM behave differently and if it can execute things that can be disastrous. So these are angles we’re currently exploring. and I have a talk about this here at Voxay Amsterdam as well. So yeah, that’s that’s that’s my interesting research currently. My name is Anton Der and I’m community leader of Yava by the tech service of the Dutch covenant and I’m interested in everything what has to do with Yava and with the tax sermonly to containers. We have a lot of big monol lead systems and we are now all transforming them to microservices inside containers.

[25:20] Excellent. And are there two things two tips or insights that you would like to share about this? Don’t think large but think small and look what you can separate. My name is Rafael Deio. I’ve been working as a software engineer for the past seven years mostly in the Java ecosystem. But now for the past five months I started working as a developer advocate at Reddis and that’s the area of technology that I find the most interesting right now which are databases in general but most mostly real-time databases just like Reddius I think that Reddis is not just a cache it’s an actual database and it’s always has been since its beginning back in 2009 and you can actually see that when you go back to 2008 and see a blog post of Sabtorio its creator where he says that he’s actually frustrated with the database technologies that they had at the time because they didn’t allow him to scale as well web servers could scale at the time. So what he wanted was a database that was fault tolerant, highly available, scalable and that allowed you to add or remove machines dynamically with no downtime. So he wasn’t asked for

[26:20] Too much and in fact a couple of months later he came up with radius its first version 2009 that already came with persistence and master replica architecture and also different data structures that allowed developers at the time to perform operations on the side of the database that weren’t available at key value databases at the time. So one of them for example was sorted sets that was used by Reddit one of the early adopters of radius that allowed them to rank the posts as you all see today very efficiently but not only that lists in radius also allowed many companies to start using them as a message kill basically later versions of radius started adding new capabilities to radius that most people are not aware of as well. So pub sub notifications streams just like in Kafka and by 2017 we also added modules the module system that allowed developers to create their own data structures and commands and capabilities and plug this into radius in a way that they never could before.

[27:18] Of course we took advantage of that and we introduced ready search radius bloom that added capability sorry probabilistic data structures to radius. Read search added full text search capability to radius. we added JSON data structure to Reddius as well. So it started working as a document database and we also added time series data structure to Reddius so that it could work as a time series database and the cool thing is that all of that were separate modules at the time now is becoming part of the core of radius with the newest version ready 8 which is coming out this month but you can already test it which is the most efficient and performant version of radius so far. I’m Jonathan Shrunk. I work at SV Bay as a Java developer. At the moment I find really interesting the new technology from Spring Spring AI. as it’s really easy to connect your LLMs in a new way to Spring Boot. I think it’s really good to follow Christian Cholov as he is the lead developer on that project and just look for videos on it.

[28:27] It’s really cool how it works. My name is Osbom. I work as a Java freelance software developer and well I use I work at the Java back ends mainly Spring Spring Boot. I’ve been working a bit a lot on GPA so persistence and lately more on architecture things. well things like encapsulation with spring modulates and I study a bit in that a bit more into that and now I talk about that and are there one or two things you’d like to share some insights that you have that people might not know about well the value of encapsulation is well I would study that it’s it’s well highly helpful in your application like well Java offers us encapsulation just with package private but if you want to go up well things like spring modulate it could help. I’m Sush. I’m I’m a solution architect, software developer. Yeah, something like that. But currently, I’m really focused on the software security especially software supply chain risk in software security. You should first update your components and dependencies because there’s a famous saying that says that you will die as a hero or you live

[29:42] Long yourself to see yourself as a villain. So it what happened to Java 8? So it was a really nice thing but the people forgot to update it and now Java 8 is also a villain. So also with other dependencies that you have in your project. So please update them with renovate depend about the snake nexus IQ whatever but just update the dependencies because there are vulnerabilities or I don’t know problems in those dependencies that maybe cause some issue for you in the future. So this is the number one recommendation and the second one is try to use dependency scanning tools before you go to production because then before you go to production you will be well aware of the components that you used in your application and the vulnerabilities that they have. So those things I think are really important updates and also use the dependency scanning. I’m Artam. I’m a principal engineer at code nomads and I think especially telling from the talks that we have today at this conference one of the big things these days is applied AI. So what I like about it is that it’s not just an abstract topic

[30:50] Because there was a lot of hype in the recent years about AI but there was a bit of a solution in search of a problem. So there was no real application of it in my opinion that was worthy especially because not the projects that I’m working with at least there’s very little let’s say opportunity to apply AI but these days with the growing precision of AI but also with the technologies around AI like the different libraries like spring AI which is presented today LangChain4j these libraries are providing very nice tools that can actually help to develop working solutions and something useful with AI and I’m I think this will only improve further. So then we will see interesting products and projects using AI that will be useful for us not only as developers but also in the products that we are making. Hi, I’m Ka. I work as a Java developer and I think there’s so much comes to mind. So I’m thinking what excites me is how much you can learn always continue to learn. You see also here at Fox days there’s so much to learn and the enthusiasm of everyone in

[32:03] This community. So when I first started to go to conference I didn’t really know what to expect. but for me it’s what excites everyone thinks about the sessions and the content you learn. But always amazing to see how enthusiastic everyone is and to speak all the other passionate people in the hallway. What do you think it is that makes people so passionate about what they’re doing? I think there’s a combination and it differs for people per person. So some people are really passionate about the details and they really want to a bit also the stubborn nurse like the people really want to figure things out and then they just continue and are so passionate about finding it out and but others I think it’s more the learning that they want to keep learning. and also maybe a bit like the how do you say the geekness or interested in small details and then just diving into it. I also sometimes think that it’s the most passionate people in a company who are developers who are the ones coming to the conferences and the ones who are less passionate are the ones that are not coming. And so you end up with a

[33:13] With a collection of very passionate people all making each other enthusiastic because because they are naturally passionate people. Yeah, I think that’s also partly true because if I look at the company when I mention like oh who goes to the conference and then you have the people are like yeah I go people are like oh unfortunately I can’t but I want to and people like I’ve never heard of it. I’m like how can you have never heard of it? And then I noticed, okay, so I’m in a bubble of people loving it and enjoying it and the meetups and there’s people what’s also fine just more do the regular job. so yeah, I think that’s that’s also true and that’s maybe also the good part about the conferences that you have them all together instead of that you have to find them in your company. Some bubbles are good bubbles. Yeah, that’s true. And but also I think that sometimes it can help if we try to get people to come also with us to conferences or meet up because I have colleagues where I think they would love it but they’re a bit like yeah I don’t know what to expect or how does it work or it will be

[34:18] Too difficult it it’ll be too hard and I won’t understand the speakers and yeah but otherwise I can’t just work and we have deadlines. I’m like yeah but this one day will not be like whether you make or make not your deadline. Yeah. So what is the message out of all of this? come to the meetups, come to the conferences or maybe at least or even if you do it within your company like find each other, go share knowledge, share enthusiasm because I noticed that when I have also people just regular days at work if we more shared enthusiasm like I also get happier and want to do things better and continue. And also now because I was at a session with IoT and then I like oh maybe I should like dust off my Raspberry Pi which is like far away somewhere and I know that in the end on a day like this I have 20 ideas and I probably won’t do most of them but that’s also fine. It helps to light up it a little bit that yeah that I do love this profession. I am Eddie Foss. I’m the chairman of the DEX for kids foundation in the Netherlands.

[35:32] DEX for Kids Foundation is for volunteer IT people that yeah voluntary give free lessons to children between 8 and 40 years and when we do that at primary schools and also at companies. The past year we developed an platform called little demand response website for schools. So they can register see if there is a volunteer teacher in the neighborhood with specific modules and then they can request help from the from that volunteer teacher and they that one will go to the school and give out workshops. That’s amazing. So who are the who are the teachers or who are the volunteers the mentors that are available to teachers? is mostly people from the open source community. So we have Java developers, Cotlin developers, u NodeJS developers everywhere. Everybody who likes to teach coding to children is volunteer at our side. So and what age group are these children? We target between eight and 14. So because with eight they have an a good understanding what we are meaning with it and processes everything also they can yeah keep good attention at the edge and we

[36:50] Target to at least 14 because at 14 they are at the moment at school they have to decide okay which study do I go to and then we finally hope to go to do a beta study so going into science so go to action. someone listening to this who wants to contribute, who wants to be available to schools as a teacher or as a mentor, what should they do? They have to go to the website little.org or go to the defox forkids.nl website and register as a volunteer or contact us by email or on our website. And are the engagements valuable? Do you get good positive feedback? Do the volunteers who do the teaching, do they enjoy it? What’s the experience like? They really like it. Also the response from the kids because there’s Yeah. They really like how to work with coding and especially if you bring also your robots with us. We have robots and you can loan them from us to use them at the school at and yeah how more they do how more they see the happy kids. So yeah, more often they volunteer to let’s do another session.

[38:00] Well, my name is P from Meova. and I’m currently very interested in ways to improve the developer experience around code reviews. reducing the well reducing the load of the review. trying to get away simplify or say take away any distractions that you usually have like minor issues and code requests or code pull requests that you can automatically fix. There’s two particular tools I’ve mostly learned about. I think the first one that I really like, I learned it at picnic. it was errorprone and there’s also picnics errorrone support which is sort of a it get tells you at compile time things that could be improved just like check stall and sonar for example can also do but in contrary to those it also suggests fixes to automatically be applied so that you can automatically sort of easily migrate your codebase as well to the new sort of standard that you want to set and the other one which is sort of in a similar place it’s it’s open rewrite which is a great tool for doing mass migrations And this decreases the load and reviews is that you do a migration all at once. What you

[39:02] Often see is that you move to a newer Java version for example and then you don’t migrate the whole code base at once. So every time you touch something you improve it a bit. But that means that you add and introduce a lot of noise on all these individual pull requests rather than just creating one big to for example migrate from JUnit 4 to JUnit 5 to use the latest Java constructs or to maybe want to move to user J or make them all slightly better. So I think those those are two topics that I’m still working on and introducing and improving. My name is Hano Emres. I’m a teacher and technology advocate at Info Support and one of the things I get the most energy of is teaching people about technology. So teaching about Java enterprise application development and also how to use AI and Genai in a good way. We live in this era era of AI right and the developments are going very rapidly. and the thing that always bugs me then is can we still think about how AI makes us better developers because I don’t want AI to take over the work that I find the most interesting right or that I that gets the most fun for me. So

[40:07] What I try to teach the people that I also teach about Java is you know use Genai in places where you think it is fitting. but always read the stuff that is generated back and have an opinion about that. And also when you’re trying to learn a new technology, it’s far better to dive into it and deal with the problems yourself than just to ask an AI agent to generate the entire solution for you because that will get you a good results in the short term probably but that will not help you in the long term when you run into more problems. So I think that’s it. So the idea to even though you can use AI for everything to be selective about where you use it because it prevents you from learning the things that AI does automatically. Yeah. Yeah. And a good example is I think you see all these social media accounts that use generative AI to generate art or music.

[41:03] And I kind of feel mixed about that because that’s also stuff that people like to do, right? We like to create art and music and we shouldn’t we should just let AI do the boring parts for us and not the fun parts. My name is Martine Vano. I work as a solution architect at Topesk. I also give trainings and one thing that I really like is when training or helping other developers to learn to come come up with creative ways to help them. And I think gamification is something that can really help in education in training people and helping them to learn new things. So in what context have you done this and what are some insights that you’ve gained from this approach? yeah so I’ve done several workshops at Tesk. One of them was about internationalization and I created some some some some puzzles some online puzzles internationalization puzzles. and I was really trying to engage people to learn something about the topic that most programmers find kind of annoying and frustrating. to approach it in a fun way. And

[42:10] So I actually organized an event where I put all these puzzles online and people tried them out and it was really cool to see people really engaging with it, sometimes even being very competitive, trying to solve the puzzle as quickly as possible. and so yeah, this is what I like to see is yeah, a little bit of gamification helping people to really get engaged and enthusiastic and learn something even if it seems difficult or impossible. Perfect. What is one thing in the internationalization context that you learned or that you found surprising or that is interesting to talk about? I learned a lot myself while making all of these puzzles and it ranges from learning about UTF8, how it works, what you can do with it, the crazy stuff that is in uni code. maybe one thing to highlight is so uniode is so complicated that there’s also even security problems that can surface. For example, and there’s one puzzle about this. If you look at right to left text supports then sometimes that means that using the right right to left text

[43:25] Markers you can actually invert parts of your code and then it looks like different code and so that could be a security problem if the code looks like it does one thing but it’s interpreted a different way because the interpreter doesn’t adhere to the right to left embedding markers and so I created a puzzle to about that but it’s also I think an important thing to be aware of that these security problems exist under the hood if you follow the uni code standard. My name is Sh Fushing and I’m Java dev but also DevOps. Let’s I try to be a bit of a allrounder and what inspires me is just the innovation and the way that people solve problems. today at the at this conference it’s been a lot about AI but then also are we going to lose our jobs to AI? And are we I would say no. TLDDR short answer no. it’s just another tool in the tool set that you have to think about how to use it and when to use it, when not to overuse it. One of the guys today said that it’s like a hammer. You know, we got a we got a tool and there’s this old saying that, you know, if all you have is a hammer,

[44:50] Yeah, everything looks like a nail. And it reminded me of the children whenever they get a new toy or a stuffed animal, it has to go everywhere with them. So it feels like that with AI now or let’s me say specifically Geni where it’s just everywhere now and everybody wants to use use it for everything and yeah so the creativity there is sometimes inspiring sometimes a bit scary. So what are one or two interesting applications of that or what have you been doing with it yourself so far? What are your insights around this? Okay. So first what I’ve been doing with it is also trying to see if we can have a conversation trying to use it like a sparring partner. not just give me answers, you know, do my job for me. trying to think of different angles which I haven’t considered.

[45:50] So I have this problem let me explain the problem to it and then can you respond in let’s say a Socratic manner to ask me yeah have you considered this have you considered that so or I don’t like this answer can you reframe it so almost like a outsourcing some of my critical thinking to it and if you frame it correctly I think yeah it could help a lot with offloading our thinking and not necessarily the coding or let it code for us. So yeah, that’s maybe a different way to think about it. And then the second part was to not be scared to go back to declarative Java because now we also saw a talk earlier today about using AI and help it to write all this streaming inputs for you and you know make it look very fancy but things get complicated and one of the speakers reminded us that code gets read a lot more than it gets written. So maybe you think about that how easy it is for your mentally unstable colleagues to read your code. So that’s joke obviously. but then not to be afraid to go back to all the tools that are out there to not focus just on one tool because then

[47:22] Java is now 30 years old has a lot of backwards compatibility but it hasn’t stopped growing. It is yeah it’s just added to the toolbox. So if you think about okay how can we keep this simple and concise it’s about yeah efficiency of code and code readability I feel at the end of the day I’m Yus I work as a CTO for driver Amsterdam and I’ve been working with spring source and I’ve been in the Java space for 25 years already and actually a lot of what we do is building custom enterprise applications so on the one hand I’m looking for information about practical stuff I saw some interesting things about extending security open connect effective containerization of applications but I’m also learning about a lot of new stuff like currently the AI talk from Christian Solof that I’m currently attending actually because we have a short break now and what are one or two insights or tips that you have gained well on the AI side definitely that this is I think the time for Java developers to start experimenting and investigating with this technology not just for building applications as a

[48:32] Programming assistance but actually to integrate in the applications themselves. so that’s definitely one and yeah on the other hand I think attending conferences not just for the technology but also for reconnecting with people again is has been a great experience so far today. Thank you Han for being the podcast interviewer and to all the guests who shared their experiences. Thank you for listening. Please subscribe to the Foojay podcast in your favorite app or on YouTube. Keep an eye on Foojay for future articles and podcasts about development and everything related to the Java world. See you next time. Give me a give me a J. Give me the friends of OpenJDK.