How a Group of Developers Took Back Control from Enterprise Java | Spring: The Documentary

This is in some breaking news. Rod, I hear that you have a big announcement to make. [music] Yes. Today we're announcing announcing announcing. When Java first appeared in the mid '90s, it was in many ways a breath of fresh air. It was a simpler language and it made a lot of sense to developers. And things started really promisingly with the serlet API getting uptake to build highly scalable websites. 1990s you had the invention of the J2E Java 2 Enterprise Edition spec and that added a tremendous amount of complexity that frankly was alien to the language itself. So by the end of the '90s we were starting to experience this explosion of complexity that took the beautiful simplicity of the language and resulted in essentially a morass that made people very unproductive. And I remember looking at hundreds, in fact thousands of lines of code one afternoon and realizing that almost every single catch block actually was more errorprone than any conceivable alternative. So you know it was those moments when I looked at the actual code and thought this really isn't matching the vision. I was writing a book that I decided to write which essentially was going to describe how J2E was great and how you just needed to suck up all the complexity. It was almost like a priest writing a book about their religion. As it went on, it took a life of its own and people approached me wanting to use that code in production. And the forum around my book was pretty active. Probably we're talking late 2002, early 2003. And one of the most active people was Yogenhura. Everything he said made complete sense. So in December 2002, I picked up a book called Expert Oneonone J2E design and development written by Rod Johnson. I read the book over Christmas break and uh I really liked what I was reading. I experienced many of the problems um firsthand before. So I I I kind of knew what it what he was talking about. that connected to everything he was writing about. Um, I was saying yes, I'm up for it. Uh, let's set up an open source project on Source Forge. Um, which we ended up doing in February 2003, not much later. Um, the hardest part actually being the name. We had to choose a name. In the early days, I wasn't quite sure um how far we could take this. At that point in time, in the early 2000s, there were very large companies dominating the space. [music] So you had Sun, you had IBM, you had BA systems, you had Oracle, you also had more peripheral uh players in the enterprise Java space who were nevertheless enormous companies like SAP. From say 2003 through maybe 2008, there was an intense battle that was waged by people [music] of goodwill and people who had rather more um debatable objectives. I was hired in 2010 as the first Spring developer advocate and uh my goal is to help people see [music] how to build interesting things with the Spring uh ecosystem. Spring was originally born, ironically, in a world where there was a group of people that were claiming that everything needed to be standardized [music] for it to be useful, which is clearly not the case today, right? Nothing. I can think of no major ecosystem where [music] all the interesting stuff is being done by standards, right? In the JavaScript community, in the net community, in the PHP community, in the Java community, all the interesting stuff is just [music] people building really good solutions and open sourcing it and giving it away and letting people build on that, [music] right? So that was the original context. Spring came out and said, "No, we're not going to try and standardize this. That would be too slow and the results would be uh insufficient." The developers were really handed down. It was a bit like, you know, handing down the tablets that will tell the developers this is the law. Really had almost no input. You know, there are a few things that make any open-source developer happy happier than seeing one user help out other users. [music] There were a lot of existing vendors with existing stakes um in the infrastructure [music] business at the time. Uh we also were disruptive to them in a way because previously they they sort of owned the programming model that came along with their infrastructure when there really in the beginning wasn't a lot of acceptance for it. uh there were big companies promoting you know their big servers so they can make a lot of money [music] and you know here's this you know young guy on a panel with all these major companies saying ah you're kind of doing it wrong let me show you how to do it right the concepts that spring framework brought to the table like dependency injection test-driven development aspectoriented programming combined sort of with the agile software movement at the time it's really groundbreaking and you know Rod and Jurgen were you know incredible visionaries and leaders [music] to promote this the direction that we took from there uh where we really took over the enterprise Java landscape uh in being the definitive provider of a comprehensive programming model for enterprise Java. I think we were part of a wave where developers took back [music] control through open source. They actually started to produce things that really worked [music] in a grassroots manner. we were getting a passionate community coming in and using and contributing and discussing and I think I can't remember when it occurred but when we got to our millionth download I was pretty convinced this [music] uh was a real thing. 2011 when I joined the spring framework team things were quite different because we were already part of a very large organization. Uh so now we had gone from 30 people to thousands uh tens of thousands of people in one company. So I moved into the spring framework team to help expand uh web development and at the time there were only two people on the spring framework team [music] uh one of them being Jurgen. Um, so for me that was an incredible opportunity to work directly. Uh, maybe even a little bit scary but definitely a fantastic opportunity. When I think about the culture of the core spring framework, the first word that comes is [music] respect. Um, that's something that I felt early on. It's often said that culture comes top down with um Jurgen um always setting a very respectful tone really respecting the community [music] in the opinion and I first felt it in the forums of the spring framework being a user and that [music] is just throughout the community. When I first started using Spring, it was in 2006, I [music] think. I had a friend who told me, you know, you you should really look into that Spring thing. [music] It's really interesting. And we were like free to choose whatever tool we wanted. I started to look into it and it was quite interesting. There were like really two things that felt really different from the get- go. Uh the first [music] one is testability. With Spring, it's very easy to craft your own application context with only the the beans that you want and run that in in [music] a unit test. You could just use them on your machine. You were not like uh dependent on an application server or something external to provide those features. Then the other thing that felt very different how easily I could find information on the web. So I would search I had I was stuck and I would search spring something. So I would get like 20 hits with like people trying to community helping each other adding adding answers editing answers. It was awesome because I could very easily find find those and that that's where really when I joined that community and I start really contributing myself once I knew what I was talking about. It's always important to consider the end user in mind and try to put yourself in that person's position. So, I think a really important skill when you're building a framework is to try to read the code you're writing in your examples. Try to read the code that you see users writing in their own applications and think, can this be simpler? You know, you've got to be relentless in thinking, okay, that doesn't look quite right. Uh, how can we simplify it and improve it? Building the business was quite challenging at the beginning because none of us had really much business experience. And so there was a whole lot of learning around okay, how do we monetize this? Do we sell consulting? Do we sell training? Do we try to become a product company? And eventually what we learned was that the only thing that could scale to the point where we could make the investment in the software that we felt we needed to make and I'm talking the open source software. The only thing that would work would be genuinely being a product company where we had the ability to raise money and the hope of achieving more significant and more scalable revenue. I never saw this as any conflict between the open-source and our journey as a company because you know the the biggest thing all along was we want spring to be the best it can possibly be. So given that everything we did really rested on the software being extremely good, there really was never any conflict. It was very clear as we succeeded in first raising money and then making money that that money much of that money was going into paying people to work on the thing they loved and wanted to make as good as possible. So certainly, you know, from the scale perspective, it would seem ridiculous that a very small group of independent-minded people could change the direction of this ship that these large companies were directly sailing into the iceberg. But, you know, I think sometimes you've got to have self-belief. Sometimes being right, having the best technology does win. We heavily disrupted that space. We we basically came in wanting to own the programming model part of it, the developer experience part of it on top of the platform infrastructure that we were running on. We effectively separated the notion of the programming model, the developer experience from the fundamental platform services from reactive programming to AI, batch processing, security, data integration, microservices, [music] a framework in the Spring ecosystem that will help you there. The reason we're able to do that is because we have a solid component model at the foundation and then as we get these new verticals, we're able to build these new frameworks on top of that. Our focus was always on the developer. Spring as a framework was trying to address uh the developer view of things, the developer experience as we call it today, allowing for the developer to do things the way they wanted. In the following years, uh, we started creating additional open source projects on top of the core spring framework such as Spring Security, Spring integration, Spring Data, setting the path for what became Spring Boot. Spring Boot was really a renewal of Spring like it was a real second [music] wave. Uh, and of course this was after I had left. So, you [music] know, it was people like Dave Sire and Phil Webb who led that charge. So, I can't directly take a lot of credit for it. My name is Dave [music] Sia. I work on the Spring Engineering team. So, Spring Boot kind [music] of ties everything together. sort of embodies in actual code what the connections are between all the different modules in Spring because it's kind of an umbrella. Spring Boot provides an integrated starting point experience uh bringing all of those portfolio projects into a preconfigured arrangement, a flexible preconfigured arrangement for getting started uh but also for just convenient availability of all those services at runtime. [music] and uh Spring Boot now is the the the preferred way of uh of of consuming the Spring Portfolio. So it became very popular very quickly. User feedback comes um very importantly in the form of um suggestions and ideas. Um probably the best example for [music] that I can give is um spring one conference maybe in 2012. Um, we had those birds of a feather sessions which are in essentially um, essentially Q&A sessions. Um, and I remember the doors to the rooms were closed and so we had to have the session um, in the hallway and everybody just gathered around and we started talking. When we got this issue asking about support for running Spring inside of an embedded web server. Well, that issue seemed pretty innocuous and trivial at the time would blossom into what is now known as Spring Boot. Essentially what they what the community wanted is I want to craft a main method and run the main method and [music] it will just start the app. Before Spring Boot you had the core container, you had like uh [music] different Spring projects like Spring Data, Spring integration that really really meant on a specific topic but you had nothing that would gather them to build an application [music] and Spring Boot is really doing that. Spring never had an application as a top level concept. [music] There's a relationship between modules. You've got framework at the bottom. You've got Spring Boot on the top and then you've got like a um sandwich [music] shape with all the little modules in the middle. So, Spring Batch, Spring Data, etc. They're they're in the middle layer there. Spring security is a [music] good a good example. What I think made Spring Boot so successful was that it did something [music] Spring has always done. Spring has always adopted good ideas regardless of whether we came up with them or whether they came from somewhere else. I started working with Spring in 2005 and the thing that drew me to it really was the readability of the fact that I could actually download the code and read it. Spring Batch is my brainchild I guess originally. So what's batch processing? It's when you you you you have a bunch of data. You have to process it from one form into another. [music] You're generally doing it offline. And so you want to have a single tool. End of day processing. You've got your invoices that you need to process. Book your income. You can't do that unless you've analyzed all your invoices, processed them into some ledger system. If you don't get that done, you can't book the income. So, we don't care who sends us the message. I I'm just going to do some work. I'm going to read some data and then write it over there. I don't care where the message comes from as long as it comes. My role have been historically to introduce new technologies in the Spring ecosystem. at least be at the forefront of it like gravlin n safety annotation and share that with the team to embrace those technologies [music] and so it's not something that is driven by some chief or CTO it's really about being empowered as a spring for committer to introduce those technologies [music] with the validation of the team cotlin is important for the modern development experience with spring because it allow to create very efficiently uh program with a very expressive language [music] Cotlin had an influence on Java and Spring design. For example, the new safety API that we now support are directly inspired from Cotlin which support those [music] features natively. And so we try to provide those features to use a huge amount of Java developers in order to bring the same benefits. [music] And hopefully I hope also that we will be able to promote more cutly multiplatform support in order to share some validation logic business [music] rules template rendering between both front end and back end in order to basically put together uh the power of spring on the server side the versatility the versatility of Cotlin and basically put the two together. I was at Cotton Conf yesterday and Jet Brains announced the strategic partnership that we are starting between the spring and the jet brain team. You know, IntelliJS and Spring's story goes back 15 years now or even longer. Meaning that back in the day 2005, there was no Spring Boot and we already supported Spring back in the day. [music] So simple spring framework functionality we already tried to make life easier. You had XML files. You could navigate around those XML files. [music] These are you know the long lost days. And then Spring actually started you know developing itself. Spring Boot came around and a ton of other you know Spring projects along the way. And we always try to be part of that uh evolution of Spring. And I think that work essentially over the last 10 years or 15 years really showed where we just try to every new feature that they had, we try to follow up and you know give you give you the nicest support for for I'm Marit Fond. I'm a developer advocate at Jet Brains. Jet Brains helps to contribute to Spring and the ecosystem by collaborating with the Spring team, making Cotlin a first class citizen in Spring, which is a collaboration between the Spring team and our Cotlin team. We have dedicated Slack channels with some of the Spring team members and bi-weekly meetings where we discuss both the new features or features that are new in Spring and show them also demos of what we are building in order to support those features so that they can give feedback to us. But also we can give feedback on the features how easy they are to use. Sometimes we find inconsistencies that we can discuss with them. Apart from that we're also part of a group that are working on JS specify for nullability support along with industry leaders. So helping to shape uh the ecosystem. You know our function as intellig idea is to be sort of a bridge between users and [music] the spring developers and the framework itself. So we take all the feedback from users which they give us on at conferences in emails issue tracker so and so forth. We take that feedback we have a look at what the spring people do. They essentially lead the entire ecosystem. They build the new features. But we have to bridge whatever they build with what the users want and bundle it in the smoothest experience possible into Intellia Idea. Developer happiness is really important for us. We want you to have the best experience possible whenever you work on your spring boot application. We don't want you to be frustrated whenever you debug an application for example. And that's why we're trying to make the experience as smooth as possible [music] in your day-to-day work life. The Spring Debugger can help you by visualizing which beans are loaded, which properties are loaded, and where they are coming from. So to help you understand the state of your current application. [music] So if you've ever been confused by uh Spring, the Spring Debugger can help you to understand that code. I think it's really fantastic to see the collaboration between Jet Brains and the Spring team [music] and community. I think the support in Intelligj for Spring is really excellent and it is [music] literally getting better month by month. By the way, I think one of the major reasons for Java success is the quality of tooling. I think great tooling is essential for a framework success. Essentially, you want to get started easily. you want to be able to work with a framework easily and that you can only get through proper tool support. So I think the tool benefits from the framework and vice versa the framework benefits from the tool. So what we're actually trying to do with intellig idea is give you the fully comforting tool available to serve all your needs whenever you're working with a spring web application be it a command line application it doesn't really matter whatever application you build as soon as you have a spring inside um we're trying to give you the best support. I think that's a great win-win for the community. It's really good to see that degree of engagement. And this is actually something that I think is really great about the Java world in the 2020s. In the 2000s, like we literally wouldn't have got the time of day from sun, even when spring was extremely popular. So when I was looking at making my last job change, I wanted to work for a company that was heavily involved in open source software development. And I learned about Hero Devs and I learned that they're definitely committed to open- source projects. So, while software developers are hopefully using the latest versions of Spring and the newer projects in the Spring portfolio and keeping things moving forward, they're also going to have those old applications that are sitting around that don't get touched very much anymore. And if there are security issues in those, it might be easier just to keep using the old versions of the projects, but fix the security problems. And that's where Hero Devs comes in. [music] We can step in and help out uh to keep those old applications safe while you're working on the new stuff. My name is Steve Paul. I'm a developer advocate who works for a company called Heradevs and they're a company that provides end-of life support for open-source software that no longer has support. My history is that for most of the Java lifetime, I've been on what would be the other side. I'm a JVM guy. I work on the runtimes that that Spring sits on and other things. So for me it was a natural and transition to move from having a conversation about [music] all the bad things that could happen to you theoretically to actually working for a company that is actively trying to fix that problem. The challenges for um an organization that's reading running older software that your team will not be learning the new things. So that's a challenge. But a lot of the the bigger challenges is that simply software though you may not realize it. Older software gets [music] um security vulnerabilities. So if you're frozen on an old version of Spring Boot, you're also going to be frozen on older versions of all those other libraries just increasing the chances that something in your stack is going to have a security problem. One interesting thing that we discovered when we started looking at some of the older projects uh was with Apache Tomcat and Spring. So you couldn't use the the last version Tomcat 8.5 with Spring Boot 1.5 and Spring Framework 4.3. So then we took on to Tomcat 8.5 as a never- ending support project and started fixing CVS and Tomcat 8.5, but it wouldn't work with Spring. So now if you are stuck on those old versions, I would never advocate that somebody sticks on that older version of projects from almost 10 years ago. Hero dev's job is to look at the older technologies and say can we provide support for them. So they would expect to see more and more of that as software [music] moves on as the head moves on as the developers move from this release to the next release [music] as their eye moves on to the next thing. Hero devs will be there to pick up the pieces behind and will provide support for those older releases. Yeah. So, Spring team has always been at the forefront of innovation on Java from the beginning and even recently with uh new developments like grow VM native and GraphQL Java app class data sharing all those new things coming up in the Java world. Uh Spring works closely with those other teams to keep the whole ecosystem moving forward. [music] having seen what the spring guys have done over the years is that uh they'll rise to that challenge. They're already producing um technologies in this space and so I don't see that being particular challenge [music] and from a hero dev's point of view the interesting thing is that [music] there's always going to be legacy software. There's always going to be software that goes out of support [music] and with what we see with AI is actually pick software components that are older that are out of support. So I'm pretty certain the hero devs are also going to be being um used by companies who discover that unpleasant truth that the AIS have built stacks of software which they need to maintain but the software in those stacks are older than um perhaps what you normal developers humans would have created. The Spring AI project is a good example of the Spring team looking at the whole software development ecosystem and trying to look for ways to help developers. And that project was started pretty recently as a way to try to help developers to uh build AIcentric applications. Spring AI came about uh by actually a podcast I listened to where they suggested uh to learn more about AI technologies. uh as a general developer skill and at the time I was working on the spring CLI project and [music] always been interested in code generation and I decided to see using the spring CLI project if I can connect to a large language model ask it to generate code and for me that was very early aha moment that wa this thing is really working I actually got it to generate code and incorporated into an existing project and I thought wow this is really powerful really felt this was the next big wave uh of opportunity for for developers [music] to create new types of applications. So, Spring AI fits into the developer broader workflow by uh providing a new API to do AI interactions. However, it looks very similar to other APIs. So, your ability to switch to this new model or this new programming paradigm for AI should should be easy. And we provide starters, spring boot starters, and those are the ways to for example quickly create tools that are exposed to the model context protocol. So with very little effort, all of a sudden your code is now part of this larger ecosystem from this model context protocol standard that is very [music] popular. The sophistication of Spring as a technology enables us to build better Genai technologies than exist on other platforms. This is where InBable my new framework builds on top of Spring [music] to provide a programming model that I think is actually quite a lot more sophisticated than you see in Python frameworks. I think that while many people still have the misconception that [music] Python is the natural language of Gen AI, I think Java is the natural language for a whole lot of Gen AI, [music] at least in enterprise. And I think that's really good news for Java as a whole. And I think Spring will lead the way as it's done for many years. Right now we are in this amazing moment. the concepts of traditional server side engineering uh and especially those with Java [music] are being challenged right to be able to wield that to take advantage of that from your software to add that kind of intelligence that's a whole new paradigm there just it just wasn't possible before right and this is happening around us so [music] Java and Spring exist at the center of all these new opportunities and uh you know hopefully hopefully that's something that people can take advantage of so with the rise of AI and other technologies including Cotlin there's a line being drawn between what you ask AI to do and what you ask and the uh infrastructure they interact with. You certainly are not going to be want to be in a position where you're going to ask an AI to have complete and unfettered access to your bank account. That would be not a good idea. But you might want to build new APIs to your bank account that your AI can make better use of. And so Spring and all the other technologies that are non AI, their challenge for the future is to adapt to this new world. I've been involved in many projects and this is the one that has had the most contributions from people in the community. It's really impressive and it's very humbling to see, you know, all these smart people all across the globe, over 250 people by my account, you know, from small documentation changes to full feature sets. Uh very polite, very collaborative environment. [music] Uh it's really a joy to work with those kind of collaborators. Spring started back in 2010. [music] Uh I stopped in 2013 because I was burned out. There's so much energy that you have to put into organizing this kind of conference that uh well I stopped for a couple of years. But then uh talking with Jurgen, he was like, "Well, you need to bring a spring aio back." And if Jurgen tells me that uh uh we need to springio back, well, I do it, right? So I uh brought the spring IO to Barcelona, which is my hometown. It's more attractive for Europeans. And since then, it has organically grown to what it is today. the amount of people who come to us at the booth [music] and tell us how happy they are to be able to use Spring uh and IntelligJ together and that it helps them in their day-to-day work life and makes you know the whole experience so much smoother. It's nice to have members of the Spring team come [music] up and ask us questions about Intelligia Idea or stuff that they want support with if we are at [music] a conference. Every time Java community was struggling with something to integrate a new feature, a new [music] library, Spring could step in and help out. Spring has been very good identifying all those extra bits that developers just don't need to know about and they figured out how to hide those [music] things or or make them easier to consume. There is something unique and special about it. There are not many projects [music] that have been successful over such a long period of time. And it is easy um sometimes to think that our dominance is um established and guaranteed uh knowing uh or seeing you know Netflix and knowing that there is spring behind it online broker account or um booking airline tickets and knowing that the code you wrote is behind that. I am uh free to introduce [music] new features that I'm really keen about, very passionated about and sometimes that become features that are used by millions of developers. It's almost a surprise to me that it is, you know, 20 years later still still going so strongly. I think it's the uh the uniformity. It's the code style basically. [music] Even if you're not, you know, working with me, you're working with me. Everybody who uses [music] Spring kind of understands each other's thought waves. If I was to explain to a new developer why Spring still matters, well firstly I would point them to Spring Boot and I would say go to Spring Initializer. Create a Spring Boot application ideally with Cotlin or with Java and just look [music] at how elegant it is and then think about all the things that does at a very high level of quality. You know, this is not something that's quick and hacky. this is going to be robust. This is something you can scale up in production. So, you know, I would hope to show them that the simplicity [music] of modern spring doesn't really come at a great price. From my perspective, uh a a framework can only be successful if it co-evolves with the platform and the language that it runs on top of. These are not independent concerns. they need to work into this towards the same direction because Spring has a core component model and that component model gives you purchase in other domains. You can become a kind of a universal developer. Pick a problem that's important to you and see if you can have like an aha moment. Spring will be there for you and it will make your life easier again without too many trade-offs. So it's not like you're sacrificing power. [music] It will be there to help you in a consistent way. The real effect is the team culture behind it [music] where uh the team really uh understands the spring way of doing things in a very natural way where basically we we don't need to talk about it all that much. Everyone knows how we go about it. It starts from the top. Good people, good community. Welcome to the Monty Python Enterprise Application Cafe. What have you got for us then? Well, this is greatly appreciated and these kind of activities really strengthen the community. Fabulous.