JPoint 2015: review of applications from speakers
Hello!
Last weekend we finished accepting applications from speakers at the JPoint 2015. Among them were both interesting and not so interesting. Call for applications almost closed . Today we will accept the latest applications from you and close at midnight. If you want to speak at the JPoint - fill out an application urgently, and we will contact you within 24 hours.
')
Under the cat you will find a short review of the submitted reports and the traditional voting box.
Roman Leventov, Higher Frequency Trading - Comparison of in-memory key-value stores
Status: questionable
In this report, Roman wants to compare Hazelcast, Ehcache, Infinispan, MapDB, Chronicle Map and one-nio. The comparison will be in the plan:
Jacob Sirotkin, Sidenis - Migration from PL / SQL to Java
Status: admitted to the program
In the report, Jacob will tell about the experience of rewriting approximately 20K lines of stored procedure code from PL / SQL to Java, including the procedure for 1000 lines. It will be explained why this was necessary, how it was done and what it gave to the business.
Aleksey Shipilev, Oracle - Catechism java.lang.String
Status: admitted to the program
In this report, we look at the main flaws of working with strings: gluttony of concatenation, fornication of substrings, self-interest of internment, pride of deduplication, and so on. Find out what it costs to hopeless JDK and JVM.
Alexey Shipilev, Oracle - Squeeze me completely
Status: admitted to the program
In this report, we delve into the guts of the JVM and JDK, in search of how the platform is trying to save on memory. Let's look at the packaging of headers and fields of objects, pointer compression, reference between generations in heaps, autoboxing cache, etc.
Nikolay Garbuzov, Donriver - How we stuffed bumps and learned to bypass the rake
Status: questionable
The report is devoted both to general issues of developing a financial platform (200+ maven modules) and to some specific issues. Questions will be considered:
Arthur Pilipenko, Oracle - Closed Model in Java ME Embedded
Status: questionable
Analyzing the application code, the guys in Java ME Embedded automatically remove unused parts of the platform: methods, fields and classes of the standard library, runtime parts, and even parts of the virtual machine. The report will discuss the analysis of dependencies of Java code, reflection, native methods and optimizations used in the closed model.
Gleb Smirnov, Plumbr - Secrets - in our heads, not in the JVM
Status: admitted to the program
Using the open source code of HotSpot JVM, Gleb will sort out some seemingly mysterious problems, having plunged into the fascinating world of C-code with listeners.
Alexander Podkhalyuzin, JetBrains - Development Productivity on Scala
Status: admitted to the program
From the report you will learn about what practices and tools can be used in teaching and further programming on Scala. Scala Worksheet and its application usage will be reviewed. An overview of some common errors in Scala code will be made. In addition, possible refactorings of the Scala language, and tools for deblicating implicits will be considered. The report will be accompanied by a small introduction to the language.
Nikita Salnikov-Tarnovsky - Where is my memory, man ?!
Status: admitted to the program
The report talks about various ways to monitor memory usage by a Java application and how to distinguish between normal and potentially critical situations. We will also discuss what information and when it is necessary to gather in order to start solving problems related to the use of memory by the application. In addition, the report briefly touches the leaks of class loaders and their features.
Dmitry Chuiko, Oracle - CompletableFuture is already here
Status: admitted to the program
With the advent of CompletableFutire in Java 8, a simple exit from hell of callbacks into the world of simple and functional Java code has opened. Blocking calls are no longer needed. The CompletableFuture report with streams and lambdas will be tested in practice. In addition, the following topics will be covered:
Nikolay Alimenkov, XP Injection - Do we need JMS in the world of modern Java technologies?
Status: admitted to the program
There are many alternative solutions to organize effective message queues: AMQP, Redis, ZooKeeper, Apache Kafka, or even self-signed queues on Cassandra. Why not use them instead of JMS? This report will discuss the underlying problems of any distributed messaging system. As a result, JMS will be considered precisely from this angle.
Alexey Ragozin, Deutsche Bank - Continuous profiling of Java applications during operation
Status: admitted to the program
Is it possible to profile an application under real load during actual operation? The report will deal with the collection of telemetry applications and the subsequent analysis of this data.
Mikhail Shiryaev, Effective Control Systems - Object-Oriented Base
Status: questionable
Alternative technology EJB. Object-oriented base. Building large structures graphs trees. Uniformity, simplification and acceleration of work with associated melons. Increase software development speed.
Volker Simonis, SAP - Packed Objects, Object Layout & Value Types - a Survey
Status: admitted to the program
A report on various projects is implemented in Java SE C-like structs. Will be considered:
Alexey Zinoviev, Tamtek - Hibernate for NoSQL, well, or whatever you have in mind
Status: admitted to the program
A report on how to mapping Java objects into persistent stores. In high-load projects, every extra byte of meta-information has to be considered, and reflection can become an irrepressible waste of a wasteful JVM. Morphia for Mongo, Hector and Easy-Cassandra for Cassandra, unnamed jugglers with annotations for Hbase, Kundera for the whole row - arguments will be given for using this or that tool in different cases.
Ivan Shabalin, Colvir Software Solutions - Approach to the development of the company's service bus in the banking sector
Status: questionable
The report will describe the method of developing web services based on OSGi technology using Apache CXF, Apache Camel, Aries Blueprint libraries. Separately, development features will be noted that facilitate the upgrade of the client’s version of the service.
Ivan Shabalin, Colvir Software Solutions - Technology for generating a web service interface specification for a metamodel
Status: questionable
Report on the developed plug-in for Maven, which takes as a basis the meta-model of a web service and generates from it, on a set of XML schemas and on data from the database a unique WSDL service. A meta model is a groovy script that initializes a hierarchy of beans, describing the namespaces of a service, its operations (grouped into categories), references to service description elements from the database. WSDL is generated according to the document-literal wrapped pattern.
Igor Sukhorukov, Deutsche Bank - AspectJ Scripting
Status: questionable AspectJ Scripting allows you to collect metrics and trace interaction in a cluster in an application that consists of many distributed processes / components that interact with each other. The framework under consideration is closest in functionality to byteman and btrace, but it uses a modified AspectJ agent and its language to describe pointcut expressions.
Andrey Solntsev, Codeborne - You can also make a good API in Java
Status: questionable
A report on how to properly design an open API. Some checklist will be given: are all methods convenient, understandable, are all classes extensible, can any behavior be customized and redefined.
Roman Grebennikov, Sociohub - Scala under the hood
Status: questionable
If you do not use a simple CRUD application on Scala, but something that is even remotely connected with the load, you need to be well aware of exactly what all these monad-shmonad cost you. The report will tell you about the magic that the Scala compiler does under the hood, tells about the horrors of scala collections, recursion and pattern matching. And why is this happening.
Nikolai Ryzhikov, WaveAccess / HealthSamurai - Why should your next JVM language be clojure
Status: questionable
Clojure is a modern and substantially functional LISP dialect capable of running on JVM, CLR & JSVM, and designed specifically for concurrency. During the report, we will learn about the language of clojure, some libraries and in parallel we will create a complete interactive web application from scratch.
Vladimir Sitnikov, NetCracker - Express yourself regularly
Status: admitted to the program
The report will tell you what to do if your regular expression works 30 seconds on a line of 50 thousand characters and how to prevent StackOverflowError from RegExp at the code review stage. Approaches to bypassing the bottlenecks of the java.util.regex standard library will be shown, alternative implementations will be considered, and criteria for distinguishing a bad regular expression from a good will be given.
Andrei Kogun, CROC - Do we need a framework !? Spring Boot Development
Status: questionable
The speaker will share his opinion on how a modern framework should look and on the need to use frameworks in general. The main focus will be on the new convention over configuration framework from Pivotal - Spring Boot. The main cases of application, the experience gained during the year of application in development, advantages and disadvantages in comparison with the main competitors will be considered.
Oleg Anastasyev, Classmates - Beyond NoSQL: NewSQL at Cassandra
Status: questionable
Report on the development of its own data storage class NewSQL: DBMS, providing fault tolerance, scalability and performance of NoSQL-systems, but at the same time preserving ACID-guarantees, which are usual for classic systems. There are few working industrial systems of this class (only Google Spanner immediately comes to mind), and there are not any accessible ones. The speaker and colleagues implemented such a system in Java and launched it into commercial operation a year ago.
Anton Keks, Codeborne - Architecture of Internet Bank without Enterprise
Status: admitted to the program
The story about the architecture of the Internet bank based on the Play Framework and Plain Simple Java. It will be shown why such a choice is better, faster and cheaper than the architecture based on Enterprise solutions. The strengths and weaknesses of the Play Framework will be shown.
Anton Arkhipov, ZeroTurnaround - Javassist in the service of a Java developer
Status: admitted to the program
Report on the Javassist library for working with bytecode in Java. The main task of such tools as Javassist is the analysis and dynamic creation / modification of Java class files. The speaker will show examples of using Javassist, as close as possible to what he and his colleagues have to do in their JRebel and XRebel products to manipulate bytecode and integrate with application servers and all sorts of Java frameworks.
Andrey Pangin, Classmates - The Best Debugger - Made by Yourself
Status: admitted to the program
JDK includes a rich toolkit for writing all sorts of debugging and monitoring tools: JVMTI, Dynamic Attach Mechanism, Serviceability Agent, Instrumentation API, etc. The report will show how to use them. It will be described how the jstack, jmap and jstat are designed from the inside. During the report, another similar utility will be written. The examples are based on cases from practice, as well as real problems of developers discussed at Stack Overflow.
Sergey Kuksenko, Oracle - Iron counters on guard of performance
Status: admitted to the program
A report on what “Hardware Performance Counters” is, where to find them and how they can be used for performance analysis.
Of those who are not in this list, but who are going to come with a report - two foreign celebrities. Jacob Fain from New York will tell one of three reports:
Yevgeny Borisov will traditionally tell something about Spring and, possibly, about Groovy. In addition, we very much hope that it will be possible to come to Baruch Sadogursky . At the moment he has a banal problem - with high probability he will not have time to make a new passport for coming to us.
Dmitry Galkin - Techno-bio-beasts: the code of artificial life
Status: Keynote report. Admitted to the program
A report on contemporary art and its role in the development of technologies from early experiments with computers and robots, to the hardcore of the last decades using nano-and neurotechnologies, bio-information hybrids, new models of techno-biological life and artificial artists.
Full announcements of the reports accepted into the program can be found in the “Reports” section of the JPoint 2015 website.
What else do you want to see in the JPoint 2015 program? Write in kamentah.
At the end of the post - the traditional voting. What reports do you find most interesting?
Last weekend we finished accepting applications from speakers at the JPoint 2015. Among them were both interesting and not so interesting. Call for applications almost closed . Today we will accept the latest applications from you and close at midnight. If you want to speak at the JPoint - fill out an application urgently, and we will contact you within 24 hours.
')
Under the cat you will find a short review of the submitted reports and the traditional voting box.
Roman Leventov, Higher Frequency Trading - Comparison of in-memory key-value stores
Status: questionable
In this report, Roman wants to compare Hazelcast, Ehcache, Infinispan, MapDB, Chronicle Map and one-nio. The comparison will be in the plan:
- target features
- architecture / algorithm / implementation
- performance (time / memory) in different cases
- restrictions
Jacob Sirotkin, Sidenis - Migration from PL / SQL to Java
Status: admitted to the program
In the report, Jacob will tell about the experience of rewriting approximately 20K lines of stored procedure code from PL / SQL to Java, including the procedure for 1000 lines. It will be explained why this was necessary, how it was done and what it gave to the business.
Aleksey Shipilev, Oracle - Catechism java.lang.String
Status: admitted to the program
In this report, we look at the main flaws of working with strings: gluttony of concatenation, fornication of substrings, self-interest of internment, pride of deduplication, and so on. Find out what it costs to hopeless JDK and JVM.
Alexey Shipilev, Oracle - Squeeze me completely
Status: admitted to the program
In this report, we delve into the guts of the JVM and JDK, in search of how the platform is trying to save on memory. Let's look at the packaging of headers and fields of objects, pointer compression, reference between generations in heaps, autoboxing cache, etc.
Nikolay Garbuzov, Donriver - How we stuffed bumps and learned to bypass the rake
Status: questionable
The report is devoted both to general issues of developing a financial platform (200+ maven modules) and to some specific issues. Questions will be considered:
- development of new functionality, in parallel with the support of existing code
- integration of internal services
- integration with third-party services
- developing an automatic testing system
Arthur Pilipenko, Oracle - Closed Model in Java ME Embedded
Status: questionable
Analyzing the application code, the guys in Java ME Embedded automatically remove unused parts of the platform: methods, fields and classes of the standard library, runtime parts, and even parts of the virtual machine. The report will discuss the analysis of dependencies of Java code, reflection, native methods and optimizations used in the closed model.
Gleb Smirnov, Plumbr - Secrets - in our heads, not in the JVM
Status: admitted to the program
Using the open source code of HotSpot JVM, Gleb will sort out some seemingly mysterious problems, having plunged into the fascinating world of C-code with listeners.
Alexander Podkhalyuzin, JetBrains - Development Productivity on Scala
Status: admitted to the program
From the report you will learn about what practices and tools can be used in teaching and further programming on Scala. Scala Worksheet and its application usage will be reviewed. An overview of some common errors in Scala code will be made. In addition, possible refactorings of the Scala language, and tools for deblicating implicits will be considered. The report will be accompanied by a small introduction to the language.
Nikita Salnikov-Tarnovsky - Where is my memory, man ?!
Status: admitted to the program
The report talks about various ways to monitor memory usage by a Java application and how to distinguish between normal and potentially critical situations. We will also discuss what information and when it is necessary to gather in order to start solving problems related to the use of memory by the application. In addition, the report briefly touches the leaks of class loaders and their features.
Dmitry Chuiko, Oracle - CompletableFuture is already here
Status: admitted to the program
With the advent of CompletableFutire in Java 8, a simple exit from hell of callbacks into the world of simple and functional Java code has opened. Blocking calls are no longer needed. The CompletableFuture report with streams and lambdas will be tested in practice. In addition, the following topics will be covered:
- Moving from a locking execution based on Future to a non-locking approach based on CompletableFuture
- API parsing
- Overhead
- A simple tutorial for multi-tier systems (using a web server as an example)
Nikolay Alimenkov, XP Injection - Do we need JMS in the world of modern Java technologies?
Status: admitted to the program
There are many alternative solutions to organize effective message queues: AMQP, Redis, ZooKeeper, Apache Kafka, or even self-signed queues on Cassandra. Why not use them instead of JMS? This report will discuss the underlying problems of any distributed messaging system. As a result, JMS will be considered precisely from this angle.
Alexey Ragozin, Deutsche Bank - Continuous profiling of Java applications during operation
Status: admitted to the program
Is it possible to profile an application under real load during actual operation? The report will deal with the collection of telemetry applications and the subsequent analysis of this data.
- Monitoring and telemetry are different things, do not mix
- Selection of telemetry parameters
- Telemetry data accumulation - problems and practical experience
- Sampling stack images and analyzing them.
- Flight recorder - JVM embedded telemetry
Mikhail Shiryaev, Effective Control Systems - Object-Oriented Base
Status: questionable
Alternative technology EJB. Object-oriented base. Building large structures graphs trees. Uniformity, simplification and acceleration of work with associated melons. Increase software development speed.
Volker Simonis, SAP - Packed Objects, Object Layout & Value Types - a Survey
Status: admitted to the program
A report on various projects is implemented in Java SE C-like structs. Will be considered:
- Packed Objects - an extension implemented in the IBM Java SDK 8
- Object Layout - a layout-optimized data storage structure from Azul Systems
- Value Types - OpenJDK proposal for small immutable data.
Alexey Zinoviev, Tamtek - Hibernate for NoSQL, well, or whatever you have in mind
Status: admitted to the program
A report on how to mapping Java objects into persistent stores. In high-load projects, every extra byte of meta-information has to be considered, and reflection can become an irrepressible waste of a wasteful JVM. Morphia for Mongo, Hector and Easy-Cassandra for Cassandra, unnamed jugglers with annotations for Hbase, Kundera for the whole row - arguments will be given for using this or that tool in different cases.
Ivan Shabalin, Colvir Software Solutions - Approach to the development of the company's service bus in the banking sector
Status: questionable
The report will describe the method of developing web services based on OSGi technology using Apache CXF, Apache Camel, Aries Blueprint libraries. Separately, development features will be noted that facilitate the upgrade of the client’s version of the service.
Ivan Shabalin, Colvir Software Solutions - Technology for generating a web service interface specification for a metamodel
Status: questionable
Report on the developed plug-in for Maven, which takes as a basis the meta-model of a web service and generates from it, on a set of XML schemas and on data from the database a unique WSDL service. A meta model is a groovy script that initializes a hierarchy of beans, describing the namespaces of a service, its operations (grouped into categories), references to service description elements from the database. WSDL is generated according to the document-literal wrapped pattern.
Igor Sukhorukov, Deutsche Bank - AspectJ Scripting
Status: questionable AspectJ Scripting allows you to collect metrics and trace interaction in a cluster in an application that consists of many distributed processes / components that interact with each other. The framework under consideration is closest in functionality to byteman and btrace, but it uses a modified AspectJ agent and its language to describe pointcut expressions.
Andrey Solntsev, Codeborne - You can also make a good API in Java
Status: questionable
A report on how to properly design an open API. Some checklist will be given: are all methods convenient, understandable, are all classes extensible, can any behavior be customized and redefined.
Roman Grebennikov, Sociohub - Scala under the hood
Status: questionable
If you do not use a simple CRUD application on Scala, but something that is even remotely connected with the load, you need to be well aware of exactly what all these monad-shmonad cost you. The report will tell you about the magic that the Scala compiler does under the hood, tells about the horrors of scala collections, recursion and pattern matching. And why is this happening.
Nikolai Ryzhikov, WaveAccess / HealthSamurai - Why should your next JVM language be clojure
Status: questionable
Clojure is a modern and substantially functional LISP dialect capable of running on JVM, CLR & JSVM, and designed specifically for concurrency. During the report, we will learn about the language of clojure, some libraries and in parallel we will create a complete interactive web application from scratch.
Vladimir Sitnikov, NetCracker - Express yourself regularly
Status: admitted to the program
The report will tell you what to do if your regular expression works 30 seconds on a line of 50 thousand characters and how to prevent StackOverflowError from RegExp at the code review stage. Approaches to bypassing the bottlenecks of the java.util.regex standard library will be shown, alternative implementations will be considered, and criteria for distinguishing a bad regular expression from a good will be given.
Andrei Kogun, CROC - Do we need a framework !? Spring Boot Development
Status: questionable
The speaker will share his opinion on how a modern framework should look and on the need to use frameworks in general. The main focus will be on the new convention over configuration framework from Pivotal - Spring Boot. The main cases of application, the experience gained during the year of application in development, advantages and disadvantages in comparison with the main competitors will be considered.
Oleg Anastasyev, Classmates - Beyond NoSQL: NewSQL at Cassandra
Status: questionable
Report on the development of its own data storage class NewSQL: DBMS, providing fault tolerance, scalability and performance of NoSQL-systems, but at the same time preserving ACID-guarantees, which are usual for classic systems. There are few working industrial systems of this class (only Google Spanner immediately comes to mind), and there are not any accessible ones. The speaker and colleagues implemented such a system in Java and launched it into commercial operation a year ago.
Anton Keks, Codeborne - Architecture of Internet Bank without Enterprise
Status: admitted to the program
The story about the architecture of the Internet bank based on the Play Framework and Plain Simple Java. It will be shown why such a choice is better, faster and cheaper than the architecture based on Enterprise solutions. The strengths and weaknesses of the Play Framework will be shown.
Anton Arkhipov, ZeroTurnaround - Javassist in the service of a Java developer
Status: admitted to the program
Report on the Javassist library for working with bytecode in Java. The main task of such tools as Javassist is the analysis and dynamic creation / modification of Java class files. The speaker will show examples of using Javassist, as close as possible to what he and his colleagues have to do in their JRebel and XRebel products to manipulate bytecode and integrate with application servers and all sorts of Java frameworks.
Andrey Pangin, Classmates - The Best Debugger - Made by Yourself
Status: admitted to the program
JDK includes a rich toolkit for writing all sorts of debugging and monitoring tools: JVMTI, Dynamic Attach Mechanism, Serviceability Agent, Instrumentation API, etc. The report will show how to use them. It will be described how the jstack, jmap and jstat are designed from the inside. During the report, another similar utility will be written. The examples are based on cases from practice, as well as real problems of developers discussed at Stack Overflow.
Sergey Kuksenko, Oracle - Iron counters on guard of performance
Status: admitted to the program
A report on what “Hardware Performance Counters” is, where to find them and how they can be used for performance analysis.
Other speakers
Of those who are not in this list, but who are going to come with a report - two foreign celebrities. Jacob Fain from New York will tell one of three reports:
- RESTful services and OAUTH protocol in IoT
- Dart for Java Developers
- JavaScript for Java Developers
Yevgeny Borisov will traditionally tell something about Spring and, possibly, about Groovy. In addition, we very much hope that it will be possible to come to Baruch Sadogursky . At the moment he has a banal problem - with high probability he will not have time to make a new passport for coming to us.
Special guest
Dmitry Galkin - Techno-bio-beasts: the code of artificial life
Status: Keynote report. Admitted to the program
A report on contemporary art and its role in the development of technologies from early experiments with computers and robots, to the hardcore of the last decades using nano-and neurotechnologies, bio-information hybrids, new models of techno-biological life and artificial artists.
Sammari
Full announcements of the reports accepted into the program can be found in the “Reports” section of the JPoint 2015 website.
What else do you want to see in the JPoint 2015 program? Write in kamentah.
At the end of the post - the traditional voting. What reports do you find most interesting?
Source: https://habr.com/ru/post/251391/
All Articles