Video of the best reports of the JPoint 2015 Java Conference - Part 2

As many of you know, at the end of April in Moscow JUG.ru will hold the fourth JPoint conference. Lovers plunge into the ocean of Java-technology waiting for two exciting days with a sea of ​​communication and a bunch of reports. A month ago, I began to talk about the best reports of last year's JPoint. Today is the time of the second part.

It is important to understand that I understand as the best reports, which were at the top of our rating. After each conference, we collect feedback from the participants, calculate it and get a lot of interesting things. In particular, the sample mean by the participants for each report. The most important special effect is that, like any statistical indicator, it reflects reality only with some accuracy. The resulting error, as a rule, is larger than the gap in estimates between neighboring reports, and therefore the order of reports in the ranking is quite arbitrary . Somehow I will tell you more about this simple model from the point of view of mathematics, but for now - to hell with the statistics, let's work on the java.
')

---

5th place

Andrew apangin Pangin, Classmates: The best debugger - hand-made
Average rating: 4.51



In essence, this report is a collection of recipes for how you can explore your production. A lot of JMX, a lot of what the JVM tells about itself in Runtime, a lot about JVMTI and in general a lot of things about what you can learn about the JVM by asking about its state from the code , and not from the profiler or some other external tool .



The report has a lot of live code. I advise everyone to spend a couple of hours tonight, run any Java program (or even the Pet Clinic) on my computer or home server and use the methods described by Andrey. Believe me, you will find many interesting things!

Of the minuses - some rightly point out that Live Demo slow down the report, because time is spent on writing and running code. On the other hand, we noticed a long time ago that if examples are shown live, the people better remember what is happening. Some brave souls even try to reproduce what is happening on the stage in the air.

---

4th place

Baruh Sadogursky, Evgeniy Borisov - Epic Groovy Puzzlers - Revenge Braces
Average rating: 4.58



“Show report”, as one of the participants noted in his feedback. Baruch and Zhenya make a special format for us - puzzle players. The essence of this: the speakers play with the hall in a perverted version of the game "who wants to be a millionaire?". Some code appears on the screen (in this case, Groovy) and 4 possible answers to what this code does. Accordingly, the participants in the hall put forward versions and try to explain what is happening in the code. For the correct answers, the speakers throw gifts into the hall. Unaccustomed to - it looks very unusual.



This report is a continuation of this report from Joker 2014. This format has been around for quite a few years, which is described at the very beginning of its first part.

After watching this report, I was once again happy for Groovy and at the same time I was horrified. Cheerfully, cheerfully, helps strain your brains.

---

3rd place

Alexey Shipilev, Oracle - Squeeze me completely
Average rating: 4.62



As we already know, in Java 9 the java.lang.String class is waiting for a lot of changes. One of the main hypotheses underlying future changes is that

• strings in most applications take up most of the memory
• Most of the lines in applications all over the world are ASCII, that is, JTW UTF-16 is redundant for them

On this occasion, an idea arose - but let's save memory and store the string not in the form of char [], but in the form of byte []. The report consistently (and here we must pay tribute to the depth of study that Alexei made) brings us to some of the difficulties and problems that the JDK / JVM developers stumbled upon while making this change with strings. Many of Alexey's favorite subjects are touched - layout of objects in memory, performance testing of old and new variants of strings, low-level iron jokes and even assembler code. Cool, cool and quite difficult. To an untrained listener, what is happening will be completely incomprehensible, but people who have listened to Alexey’s previous reports will get a real high.

---

2nd place

Nikolay Alimenkov, XP Injection - Do we need JMS in the world of modern Java technologies?
Average rating: 4.63



A report on the boundaries of JMS application and its place in the modern world. JMS technology is quite old, from the nineties. JMS 1.1, which is used in most modern productions, was already written 15 years ago.



The report examines both some of the fundamental problems of JMS, and the features (mostly unpleasant) of its various implementations. Nikolay consistently states his position that JMS is not needed. The report mentions a zoo of modern technologies - Hazelcast, Erlang, ZooKeeper, Redis, RabbitMQ and Apache Kafka. In general, lovers of distributed systems is dedicated.

---

1 place

Aleksey Shipilev, Oracle - Catechism java.lang.String
Average rating: 4.74



We all have long perceived the String class as something native and simple. Here he is - warm, lamp. However, things are not so simple. This could already be understood from Alexey’s report on changes to the String in JDK9, which I mentioned above. But in the current versions of Java thongs are fraught with many surprises.



What can compile concatenation? What to do with collisions in String.hashCode ()? Where did the internal fields count and offset disappear, and most importantly, why did they disappear? How is string interning different from deduplication? How does JVM speed up the equals method for long strings? In general, a lot of useful and non-obvious.

---

What about JPoint 2016?

We are often criticized, they say, "you have the same speakers from year to year." Well, where to put these beauties, then? In my personal opinion, the fact that the best speakers come to us from year to year makes the conference only stronger. And I am very happy to note that all these guys, many of whom have not lived in Russia for a long time, despite their own workload, despite the long flights, despite the dollar exchange rate, will finally perform at JPoint 2016.

---

Links

As always, at the end of the post - advertising and usefulness. Separately, I want to draw attention to the fact that each of you still have the opportunity to become a speaker. Do you think that at JPoint 2016 some important topic is not disclosed? Write in the comments, but better - make a report on this topic!

• Video recordings of all reports of JPoint 2015 ;
• JPoint 2016 speakers (approved 13 out of 30)
• Reports JPoint 2016 (list is updated)
• last chance to be a speaker