JavaOne / Oracle Develop 2012: general impressions. Story with pictures
On April 17-18, 2012 in Moscow, in the building of the Russian Academy of Sciences, the second JavaOne / Oracle Develop conference was held. Last year I attended this event and I really liked it, so I bought an entrance ticket as soon as they went on sale. By the way, their minimum cost has doubled compared to last year, but remained quite adequate - 1,000 rubles.
Author photo - Trukhin Yuri
At the entrance I met the famous Java character Duke.
So, once inside, we were greeted by such a poster and breakfast.
Author photo - fLanKer
')
This time, the problem was almost solved with catering: the food was delicious and genuine. However, the coffee was still soluble, which I personally do not fit in my head at all: Java and surrogate coffee, brr ...
Well, okay, it's all everyday. After breakfast, I, like many other participants, went to the main hall for the opening of the conference and the plenary report in order to listen to what vendors want to convey to us. I admit honestly, I have not really departed from sleep by that moment, so I can not say what it was about. Glimpsed the words "Java SE", "Java EE", "Cloud" and stuff like that.
Nevertheless, the report ended and, according to the plan, I moved to the blue hall, where the stream “Java SE” was located.
The first report was called "JDK8 and beyond . " Actually, JDK8 and a set of its features is no secret to anyone, and I was particularly interested in "... and beyond . " The feature sheets showed so quickly that I remembered almost nothing, only in fragments: there would be no more primitives, only objects, a self-tuning JVM and many, many more. We are waiting for the slides on the conference site.
The next was the report of our compatriot Mikhail Ershov “Seven deaths will not happen, but Jigsaw will not be avoided” . Quite a lot of people came to him and in the hall it was like this:
Photo author - Oracle SPB
It was very interesting to talk about what awaits us already in JDK8, namely, about the project of application modularization. Hopefully the days of JAR Hell are already numbered :)
At the end of the speech, a dinner was held, which differed favorably from what it was last year (remember, there was a sort of camping set: sandwich, jar of lettuce, small snickers and juice): fried potatoes, rice, fish in batter, chicken skewers, pancakes , various sauces and salads. After such an abundance, people relaxed in poufs in the hall.
Author photo - fLanKer
After lunch, I looked into the blue hall and did not hear anything interesting (there was something about the Oraklovsky Coherence, but rather dry and boring), I decided to see what was on the main technical plenary report. And there they showed the general vector of development of Java as a language and as a platform in the directions of SE, EE and ME.
Author photo - Petr Popov
Author of the photo - your humble servant, Ilya Ermolov
Immediately after the big plenary report, a coffee break followed - there were tables with hot water, coffee, tea, sandwiches / buns / cakes in all corners of the territory. It was possible to try all this magnificence in the so-called “Winter Garden”, which is located in the center of the floor with small halls.
Author photo - Ruslan Pilin
After afresh coffee, I went back to the blue hall, because there were two very interesting reports planned in a row - “Methodologies for optimizing performance” and “Dragons in the household: scaling on multi-core machines”
Author photo - Petr Popov
These messages were closely connected with each other, besides the St. Petersburg guys TheShade and Walrus staged a full house - people simply could not enter the room, let alone sit down somewhere. Briefly: they talked about how to properly evaluate performance and what to do in the first place, when the application does not work as quickly as we would like.
Workshops were working throughout the day. How was it - I do not know, did not go :)
Author photo - Marcus Hirt
On this, the reports of the first day were completed, but the evening was just beginning :)
Author photo - Petr Popov
This year there was no live music. You can argue a lot here: some people like live music, others prefer a calm background easy-listening and communication (I’m the first), but in general, they drank beer normally and talked.
So, the city falls asleep, the mafia wakes up.
The second day began surprisingly the same as the first, only I came a little later, did not go to the plenary report, and immediately proceeded to my favorite blue hall. There, Vladimir Ivanov was preparing to speak with his first report, “The Garbage-First - a new garbage collector in HotSpot JVM” . In short: the new garbage collector divides memory into segments and is already working with them in incremental mode. Positioned as a replacement for CMS, although not in all aspects superior to it.
Next on the list is "Faster than Fast or Evolution of Dual-Pivot Quicksort in JDK" from Vladimir Yaroslavsky. I understood at the beginning, and then somehow I lost the thread of consciousness, because I perceive the algorithms poorly in the fast narration without stopping.
Lunch again.
After lunch, there was another great report from TheShade about the new “feature” (in quotes, because it’s rather an entire task parallelization subsystem in JDK7) “Fork / Join: implementation features, usage models, performance” . Again, it does not make sense to retell it, you can watch the slides or read the official docks.
The next was a report from Vladimir Ivanov already known to us - "Multi-level compilation in HotSpot JVM" . In short: a new JIT compilation system that combines the 2 previously used methods of client and server compilation.
After the coffee break, most of them went to the “Java Memory Model”, which they did, in my opinion, in vain. After all, a report from Cheremin on the topic “Dismembering Disruptor: Magic and High Performance Technology” began in the same blue hall. The most beautiful story about the library for a very fast work in a multi-threaded environment ( more ).
The final report was again Vladimir Ivanov entitled “Memory Management in Java: Minimizing Memory Consumption” .
Author photo - Ilya Ermolov
Basically, the story was about various types of links (strict (or hard), soft, weak and phantom), how to use them, as well as about various methods of addressing objects in memory.
The conference, as they said on each report, did not give any specific instructions for action (it should not do that, in fact). But she provided a lot of food for the mind: in which direction to develop, what to study, what subtleties to pay attention to.
Liked much more than last year. This time 5– out of 5 :)
Yes, yes, this minusik for coffee. Yandex and Google always have natural coffee, even in thermoses and already ready. But the present. Many people still go from afar, at night and in the morning really want a good coffee.
And so everything is fine! Thanks, Oracle!
More photos can be found, for example, by tag in the tweeter #javaoneru or #javaone
Some official photos from Oracle Russia
Presentations:
Seven deaths will not happen, but Jigsaw will not escape
Performance Optimization Methodologies
Dragons in the household: scaling on multi-core machines
Fork / Join: implementation features, usage patterns, performance
Java memory model
Java Memory Management: Footprint
Multi-level compilation in HotSpot JVM
The Garbage-First - a new garbage collector in HotSpot JVM
Migrating Spring Applications to Java EE 6
How to JBoss
Kotlin language for the Java platform
Dismembering Disruptor: High Performance Magic and Technology
Master class "Java EE 6 Hands-on Lab"
Author photo - Trukhin Yuri
At the entrance I met the famous Java character Duke.
Day 1
So, once inside, we were greeted by such a poster and breakfast.
Author photo - fLanKer
')
This time, the problem was almost solved with catering: the food was delicious and genuine. However, the coffee was still soluble, which I personally do not fit in my head at all: Java and surrogate coffee, brr ...
Well, okay, it's all everyday. After breakfast, I, like many other participants, went to the main hall for the opening of the conference and the plenary report in order to listen to what vendors want to convey to us. I admit honestly, I have not really departed from sleep by that moment, so I can not say what it was about. Glimpsed the words "Java SE", "Java EE", "Cloud" and stuff like that.
Nevertheless, the report ended and, according to the plan, I moved to the blue hall, where the stream “Java SE” was located.
The first report was called "JDK8 and beyond . " Actually, JDK8 and a set of its features is no secret to anyone, and I was particularly interested in "... and beyond . " The feature sheets showed so quickly that I remembered almost nothing, only in fragments: there would be no more primitives, only objects, a self-tuning JVM and many, many more. We are waiting for the slides on the conference site.
The next was the report of our compatriot Mikhail Ershov “Seven deaths will not happen, but Jigsaw will not be avoided” . Quite a lot of people came to him and in the hall it was like this:
Photo author - Oracle SPB
It was very interesting to talk about what awaits us already in JDK8, namely, about the project of application modularization. Hopefully the days of JAR Hell are already numbered :)
At the end of the speech, a dinner was held, which differed favorably from what it was last year (remember, there was a sort of camping set: sandwich, jar of lettuce, small snickers and juice): fried potatoes, rice, fish in batter, chicken skewers, pancakes , various sauces and salads. After such an abundance, people relaxed in poufs in the hall.
Author photo - fLanKer
After lunch, I looked into the blue hall and did not hear anything interesting (there was something about the Oraklovsky Coherence, but rather dry and boring), I decided to see what was on the main technical plenary report. And there they showed the general vector of development of Java as a language and as a platform in the directions of SE, EE and ME.
Author photo - Petr Popov
Author of the photo - your humble servant, Ilya Ermolov
Immediately after the big plenary report, a coffee break followed - there were tables with hot water, coffee, tea, sandwiches / buns / cakes in all corners of the territory. It was possible to try all this magnificence in the so-called “Winter Garden”, which is located in the center of the floor with small halls.
Author photo - Ruslan Pilin
After a
Author photo - Petr Popov
These messages were closely connected with each other, besides the St. Petersburg guys TheShade and Walrus staged a full house - people simply could not enter the room, let alone sit down somewhere. Briefly: they talked about how to properly evaluate performance and what to do in the first place, when the application does not work as quickly as we would like.
Workshops were working throughout the day. How was it - I do not know, did not go :)
Author photo - Marcus Hirt
On this, the reports of the first day were completed, but the evening was just beginning :)
Author photo - Petr Popov
This year there was no live music. You can argue a lot here: some people like live music, others prefer a calm background easy-listening and communication (I’m the first), but in general, they drank beer normally and talked.
Day 2
The second day began surprisingly the same as the first, only I came a little later, did not go to the plenary report, and immediately proceeded to my favorite blue hall. There, Vladimir Ivanov was preparing to speak with his first report, “The Garbage-First - a new garbage collector in HotSpot JVM” . In short: the new garbage collector divides memory into segments and is already working with them in incremental mode. Positioned as a replacement for CMS, although not in all aspects superior to it.
Next on the list is "Faster than Fast or Evolution of Dual-Pivot Quicksort in JDK" from Vladimir Yaroslavsky. I understood at the beginning, and then somehow I lost the thread of consciousness, because I perceive the algorithms poorly in the fast narration without stopping.
Lunch again.
After lunch, there was another great report from TheShade about the new “feature” (in quotes, because it’s rather an entire task parallelization subsystem in JDK7) “Fork / Join: implementation features, usage models, performance” . Again, it does not make sense to retell it, you can watch the slides or read the official docks.
The next was a report from Vladimir Ivanov already known to us - "Multi-level compilation in HotSpot JVM" . In short: a new JIT compilation system that combines the 2 previously used methods of client and server compilation.
After the coffee break, most of them went to the “Java Memory Model”, which they did, in my opinion, in vain. After all, a report from Cheremin on the topic “Dismembering Disruptor: Magic and High Performance Technology” began in the same blue hall. The most beautiful story about the library for a very fast work in a multi-threaded environment ( more ).
The final report was again Vladimir Ivanov entitled “Memory Management in Java: Minimizing Memory Consumption” .
Author photo - Ilya Ermolov
Basically, the story was about various types of links (strict (or hard), soft, weak and phantom), how to use them, as well as about various methods of addressing objects in memory.
General impressions
The conference, as they said on each report, did not give any specific instructions for action (it should not do that, in fact). But she provided a lot of food for the mind: in which direction to develop, what to study, what subtleties to pay attention to.
Liked much more than last year. This time 5– out of 5 :)
Yes, yes, this minusik for coffee. Yandex and Google always have natural coffee, even in thermoses and already ready. But the present. Many people still go from afar, at night and in the morning really want a good coffee.
And so everything is fine! Thanks, Oracle!
More photos can be found, for example, by tag in the tweeter #javaoneru or #javaone
Some official photos from Oracle Russia
Presentations:
Seven deaths will not happen, but Jigsaw will not escape
Performance Optimization Methodologies
Dragons in the household: scaling on multi-core machines
Fork / Join: implementation features, usage patterns, performance
Java memory model
Java Memory Management: Footprint
Multi-level compilation in HotSpot JVM
The Garbage-First - a new garbage collector in HotSpot JVM
Migrating Spring Applications to Java EE 6
How to JBoss
Kotlin language for the Java platform
Dismembering Disruptor: High Performance Magic and Technology
Master class "Java EE 6 Hands-on Lab"
Source: https://habr.com/ru/post/142344/
All Articles