On April 5-6, 2019, the JPoint hosted the JPoint - today, the largest Java conference in Russia (the new Joker has not yet passed). Let's remember how JPoint went through, what interesting things happened there ... and we will start preparing for Joker.

Under the cat a lot of pictures, and, accordingly, you will need a lot of traffic. Last time, habrovchane asked to pinch everything in the JPG. Now it is done, but still it turned out quite volume.

How did everything go

JPoint 2019 is currently the largest Java conference in Russia and our largest specialized conference in general. This is a very important event in the history of the Russian community of Java-developers.

Unfortunately, it was from the size that the main problem went: it became cramped at the conference. On some reports, people did not fit into the halls, and for this we offer our deepest apologies. Even in the additional hall with overhead and at the demo stage there were overbooks. There is a feeling that this is the real size of the conference, which can fit into the World Trade Center without additional tweaks .

How did it happen? The place where the conference is held, we call short - "playground". The site is chosen strongly in advance (at least one year) based on the planned load, and then, in case of large discrepancies with the plan, the decision can be adjusted. Loading halls is a key thing, we are watching it closely. Two weeks before the conference, people were not so much abundant, but even slightly less than planned - and nothing foreshadowed problems. And suddenly, at the very last moment, people abruptly and massively began to pay for new tickets. While we were closing sales, tickets were purchased by another 200 people over and above the plan. It was obviously impossible to change something. What to do next year is the question. In Moscow, there are more sites, for example, Crocus. But Crocus - it is very, very expensive, and not quite suitable for the halls and equipment. There are low ceilings, unregulated lights, echoes in the halls and something to do with soundproofing. There is Skolkovo, but this is not the case at all - all the halls are small, for 100-200 people. There, too, low ceilings - so much so that the screen is not normally hang, and also equipment issues. So yes, finding a good platform for the next JPoint will not be easy. But don't worry, we'll think of something.

The rest of the conference was a success. Even the report “The Game of God”, which is not about Java at all and therefore could cause an ambiguous reaction, turned out to be a great success. Speakers are inspired too. Some of them are discussing how to get a reusable visa to Russia (as you probably know, there are always some difficulties with visas to Russia). Someone measures how many hours he has stood in the discussion area and is going to overtake his colleagues next time. Interesting reviews from speakers that few Java conferences in the world can be compared in terms of organization and, choosing between them and JPoint, they will choose him. It seems that the emphasis on quality has begun to give long-lasting positive effects.

Photos and video reports

If you participated in the conference (it doesn’t matter - live or in the online broadcast), then all the cleaned and cut videos will appear in you very soon. But everyone else will have to wait at least six months, we plan to post entries closer to Joker 2019.

The first photos are already in our group on VKontakte . The rest of the photos will be in about ten days.

Further in this post, events are described approximately in the order in which they occurred at the conference: the opening keynout, the reports closing the keynout, bofs, and so on.

Anton Keks opening keynote “The world needs full-stack craftsmen”

This is a very unusual report, which states, among other things, that a project without at least one fullstack engineer is dead. There was an interesting moment when Anton asked: who could single-handedly rewrite the entire project from scratch if the whole team moved the truck? A forest of hands is rising. The next question is whether you will have a better project than the current one. Most likely, yes - at least there were also enough hands. But then in the course of the report it turns out that simply writing code from scratch is not enough. You need to communicate with the client directly, propose solutions, design a UI, and so on and so forth, a cloud of slides with lists. Anton continues to throw and throw skills - as a result, the discussion does not end with the end of the report - a whole crowd gathered in the discussion area, continuing the discussion of the topic.

First Day Reports

Obviously, we with the phillennium and golodnyj most of the reports could not be seen, because most of the time was spent doing interviews, communicating with companies, speakers, and so on. So, probably, a post with a review of reports would be better to write to someone else. But you, dear reader, now only I have, so I will tell only about what I had time to see!

First of all, this is a very breakthrough report on “Create your own interpreter with a JIT in 1 hour” by Charlie Gracie. By the way, how many people with the post Garbage Collection Architect do you know? I'm not sure that this post is even worth cutting as “GCA at IBM”. Imagine you are Charlie, come to the meeting and introduce yourself: “I am GCA at IBM”. And all around are: "He - who ???" In short. As expected, Charlie did not tell how to build the world in 7 days without the use of frameworks, but brought his Eclipse OMR. We write an interpreter, the interpreter becomes a compiler, in theory everything is simple, in practice - you need to understand a lot of nuances. At a minimum, C ++ is everywhere and the code you need to go into. Charlie kodit right on his MacBook, often with C ++ programming, this is a fatal error (because it takes a long time to search for bugs), but Charlie has everything going smoothly. Oh yes, this report has no slides, and how to do live decoding on Habr is not clear.

The second is a very worthy thing, "The Tale of how we did the algorithm of channels in Kotlin Coroutines" . Here the slides are already there . And you can say, Nikita is the king of slides, because everything is clear on the slides, and there is even something that was not spoken by the words live. Once I wrote a comic article about why Kotlin is bad (judging by 321 extremely serious comments, no one guessed that this was a joke). In fact, Kotlin is a fantastic thing that is created and deployed right before our eyes. When the coroutines were created, there was no support for them in runtime - and there is still no support, Pressler refuses to even hint at a release date. How does all this work in Kotlin, how can it work at all? And here comes the wizard and shows the secrets of his greatest tricks. Well, okay, with the clarification that they haven’t yet released what Nikita is talking about - it’s still excellent. Of course, it would be necessary to criticize something, and this is the question: what conclusion can an ordinary developer draw from this report? And although not a word about Kotlin in the report, for me this is a great story for broadening my horizons (Nikita says this in his introductory speech), but anyone who wants to find some skills for instant use can be deeply offended .

Thirdly, this is “Spring Boot Us, and we are getting stronger: the unbearable ease of AOT compiling Spring applications” from Nikita Lipsky ( pjBooms ). This is a very popular topic now AOT-compilation, and even the example of Spring Boot, about which intuitively it seems that AOT should not work with it. And in GraalVM it really does not work (Spring Fu does not count). And in Excelsior JET it works, and this is another great witch. As always, I see it more like a report to increase my outlook: you begin to understand better how Spring works, how Java works in general - why language and runtime developers choose such solutions, and not others. You understand the problems of modern clouds, and so on. Well, probably, if you have a purchased JET, this is also a purely working question, but I don’t have one.

All reports of the first day (with links)

Demo stage

At the demo-stage you could listen to a variety of reports. For example, here's Alexander Belokrylov alexbel from BellSoft talks about Liberica:

Workshop thomas wuerthinger

Thomas specially set aside time for us and conducted a workshop. He wanted to work with a very small audience (this turns out to be more efficient), so we didn’t make big announcements: there was a great chance that after his report the whole room would rise and move to a workshop, which is much more than the planned 20 people. As a result, everything turned out great.

BOF

BOF is now a traditional format at our conferences. Something like a round table or discussion group in which everyone can take part. This format historically goes back to the first informal discussion groups of the Internet Engineering Task Force (IETF) . There is no division into a speaker and a participant: everyone participates on a par. Nevertheless, there is a moderator on every boss, directing the discussion, and at least a few recognized experts - all together they make up the core from which the discussion begins, and then how it goes.

Microservices, cloud and where it all goes (RU)

Participants : Oleg Anastasiev, Dmitry Chuiko, Alexander Nozdrin-Carpenter, Vladimir Plizga Toparvion , Maxim Gorelikov somemaxim ;
Moderator : Kirill Tolkachev.

Reactive - today's need and future perspectives

Participants : Sergey Egorov bsideup , Oleg Dokuka, Tomasz Nurkiewicz, Vadim Tsesko;
Moderator : Anton Arkhipov.

Rumble in the jungle (Oracle JDK, your own OpenJDK build, alternatives) (EN)

Participants : Alexander Belokrylov, Simon Ritter, Bruno Borges;
Moderator : Volker Simonis.

Why does Java run slow? (En)

Participants : Nikolay Igotti, Rafael Winterhalter, Andrey Pangin apangin , Andrzej Grzesi, Vladimir Sitnikov;
Moderator : Ivan Krylov.

Interestingly, a spontaneous BOF with Anton Chernousov, Andrei Kogun, Alexei Abashev, Egor Bugaenko and others formed next to the stand of Leroy Merlin. So much warmth and positive!

A party

At the end of the first day there was a big party that allowed a bit to cool melting brains. Naturally, it is not possible to photograph the entire huge party of one and a half thousand people, so here is a motivational picture for you.

Opening keynout of the second day: Alexander Pankin, “The Game of God. Has science crossed the border? ”

This keyout is not about java at all. He said this: considering the latest achievements of genetic engineering, will we continue to play God or is it time to outplay him?

Despite the fact that the report at times concerns rather creepy things, Alexander manages to smooth corners. For example, where we are talking about "fluffy genocide." There is a famous story that there are no natural predators in Australia that can fight rabbits. And when the rabbits were delivered, they devoured everything there, there is no grass in this photo. They actively reproduce and destroy natural ecosystems. No one will give a chance to kill all these unfortunate animals, but there is a hypothetical example when special mutant rabbits can be bred, in which females will run out after a lot of crosses and they will disappear on their own. In this case, no rabbit was hurt, everyone was engaged in an exceptionally pleasant breeding process.

Second day reports

I'll tell you more about a couple of reports. It was definitely worth going to the report of Oleg Shelayev, "Supercompilation, partial evaluation, projections of Futamura and how GraalVM will save the world . " We actively discussed this topic last year in the @graalvm_ru community in Telegram, and finally, Oleg came to us with a full report. This is a topic with a keen taste of patriotism: starting to study the documentation for Truffle, you quickly come across the fact that the theory is built on the works of Russian scientists of the last century. It is not clear how the developers of Graal get along with this - most likely, they do not know how to read Russian. It's nice that nothing is forgotten, and the ideas of Novosibirsk academicians have finally acquired their engineering embodiment. And there were pineapples! (a joke for those who attended)

It was definitely worth going to “Atomic refactoring in IntelliJ IDEA: bending IDE by yourself” by Tagir Valeeva ( lany ). Firstly, because it is Tagir. Secondly, because it is something very practical. Many refactorings do equivalent transformations - but not all. If your program is not very strange, then everything is fine, but sometimes it breaks. Sometimes it breaks positively - for example, you can replace tail recursion with a cycle, while the semantics change - the program stops falling, but this is even for the better. This and many other things were in the report of Tagir, and I would review it again when videotapes appear.

More detailed analyzes, probably, will need to be done in TOP-10 or some such traditional format for Habr, where you can write an entire wall of text.

All reports of the second day (with links)

Closing keynote: Bruno Borges, "Making sense out of serverless computing"

It's unusual to see someone from Microsoft as a keynote at a Java conference, because for many years everyone got used to the fact that Microsoft is about .NET, C #, and so on. By itself, Bruno had previously worked as the Developer Advocate in Oracle. He started his Java development career in 2001, wrote applications on Swing, then on JSP and Servlets, Apache Struts ... And now he is Developer Advocate already at Microsoft. In this case, even in Microsoft, he does not use Windows, uses a Mac. We live in amazing time. The clouds play a significant role in this, and this is the whole story of Bruno. As written in one of the first slides, “Serverless is a State of Mind,” and this state implies that we can more freely choose our technologies than ever before. When Bruno started the report, he asked how many people use serverless and there were very few hands. By the end of the report, many people who had never thought about using serverless technologies had a desire to try them - this is the special magic of keynote reports when they don’t even show you any shocking low-level hardcore, kilometers of code, but simply show a big picture of everything. - and for you it becomes dear to the future.

Exhibition

The conference was held at the International Trade Center (ITC). At the entrance to the site we were waited by a beautiful exhibition, which we diligently obsutroili and filled with activities all sorts of companies. It is not possible to write about all the companies and activities, there are too many of them, therefore, as always, I’m throwing a few random photos.

At the Leroy Merlin stand, I lingered for a second in order to understand what strange iron things it was, and I was immediately offered to scan the QR code. Guess what was the link .

At the stand of Sportmaster and Raiffeisenbank they play table hockey:

This is how people discuss the details of solving problems at the stand of the Deutsche Bank TechCenter and win prizes from Mediascope:

And at the stands of Alfa-Bank and HeadHunter you can chill, chat with employees and consume a huge amount of coffee:

And it was also possible to buy a small volume of Yegor Bugayenko at the small bookshop of the Peter publishing house If, however, writing according to the precepts of this book does not work and coding is not your topic at all, then you can try yourself in Adjale:

Anything gets better if you write it on Kotlin!

Conference organization and team

I would like to say a special thank you to all colleagues who participated in the organization of the conference and who are now reading this article.

By the way, it was the first JPoint, which started 23derevo without Alexey Fyodorov, at that moment he was in Porto at the ICPC finals, but flew in on the second day and could be seen in the speaker room.

The speaker-room is generally a very interesting place, and there occur their interesting things, closed from the outside world. For example, Oleg Shelayev managed to drive his report on the projection of Futamura several times in a row and correct slides (and one of these runs was with me).

Lyosha is on the left and Oleg is on the right :)

Nobody notices photographers, but they are! Thanks to them in this post so many pictures. In truth, this is the only photograph of a photographer that I even have - it seems that they do not take pictures of themselves.

In addition, the operators of video recording and video broadcasting work quietly and imperceptibly. We have a new system for monitoring video and equipment in general, thanks to Artem Nikonov, Maxim Zverev and the rest of the participants in this feat.

In addition, software coordinators can be found at the site - thanks to them, everything works like a clock.

Finally, the halls and everything that happens there is the care of a large number of people who can solve any problems that arise. By the way, there are not many problems to say that there was a lot - once the microphone began to catch the sound from the next room, it was quickly repaired, and from serious it was all.

You can continue for a long time: to talk about our own Wi-Fi network (over which Maxim Zverev is conjuring), about the online broadcast (which the phillennium and I usually lead, olegchir ), about reports from the demo-stage ... But this post will never end.

What's next: Joker 2019

Let's summarize. JPoint 2019 is completed, it turned out to be very good, and further we are waiting for new conferences.

We are opening sales Joker , which will be held this fall, October 25-26. There are several speakers right now:

• Juergen Hoeller of Pivotal, co-founder and leader of the Spring Framework;
• Mark Heckler is also from Pivotal, the famous author and evangelist of cloud and IoT solutions;
• Stephen Chin - Director of Oracle Groundbreakers Team at Oracle, Java One conference keyner, creator of Raspberry Pi with Java, co-author of Pro JavaFX Platform, chairman of Oracle Code One Content;
• Oleg Nenashev - developer core team at CloudBees, an expert in the development of infrastructure and frameworks for Jenkins;
• Anton Arkhipov - Developer Advocate in the company JetBrains, a resident of the podcast "Debriefing".

In addition, you can submit a report yourself - the link is a good instruction on what to do.

Information about new speakers is constantly updated on the site . You can also buy tickets there .

Participants of the past Joker are strongly advised to keep track of their mailbox, where they will soon receive a letter with promotional codes.

In general, stay with us, meet at the next Joker and JPoint!