Did you ever have to cancel Open Type dialog because you didn't have time to wait for the search indices to initialize? Does slowness of populating the Call Hierarchy view annoy you? We suspect that for the majority of Eclipse Java users the answer to these questions is yes.

The proposed solution to these and many other JDT sluggishness problems is a new index containing an order of magnitude more information about Java code than the old JDT indexes and caches. The technology chosen for the new, faster JDT index is heavily influenced by the CDT index that has been in use for almost a decade and proved to be very efficient and flexible. CDT borrowed a lot of code and ideas from JDT. Now it's time to pay back.

Come to this talk to learn about the design of the new index and to see how it makes JDT blazingly fast.

Java is the language of choice in enterprises. Many of the large software systems are built using Java. The key reason for Java being language of choice was the tool support, robust JVM, statically typed language and modularity. Modularity is one of the key aspects of the design. Eclipse (built on Java) is the wonderful example of a modular software system based on OSGi specification (Equinox). Project Jigsaw which is umbrella project is aimed to design and develop a standard module system which is compatible with OSGi. Brining modularity within the platform will add many improvement and will also provide the ability to run it down to small devices. Jigsaw has been one of the major features of JDK 9.0. The talk will cover details of project Jigsaw. 

Java is mostly used presently at server side. Improvement in the server side processing and ability to evolve with the new specification and protocol would boost the performance. HTTP/2 which is the latest specification makes use of streams. Java 9 brings native support of HTTP/2.

Java has been a programming language of choice for many years in academia, till the less verbosity, better interpreters and tools for JavaScript, Python evolved. The language evolved a bit 2014 with the introduction of Lambda in Java 8, which provided less verbosity. Still the absence of Read-Eva-Print Loop (REPL) moved the academia to embrace python as the language of choice for teaching problem solving to the students. Java 9 is bringing jShell which provides REPL functionality in Java 9.

Java 9 is going to provide several developer tools which complement the modular system and enhance developer productivity.

I plan to give overview of new Java Linker tool that would assist in creating a modular Java runtime. "jlink" assembles and optimizes a set of modules and their dependencies and "jmod" for creating modular packages with class files and resources. "jimage" which is 5x more performant than zip or jar formats.

I plan to give interactive session in how to create a modular JDK image with specified root modules using jlink. Describe the plugin api for extension using custom modules.

I shall give insights on tools usage and internals which would help developers in decision making along with benchmark results.

To exploit the cross platform nature and powerfulness of the  JVM, most of the nonJava programming languages (python, ruby, clojure etc) had ported themselves to run on JVM. But performance of these nonJava languages were poor compared to their nonJVM version. Java 7 introduced a new byte code instruction invokedynamic (indy) to support dynamic languages running on the JVM. The invokedynamic instruction simplifies and potentially improves implementations of compilers and runtime systems for dynamic languages on the JVM.  Later in Java 8, indy became the backbone of the Java 8's most popular feature lambda expression.

Java Platforms team is all set for JDK9 release. I am associated with G1 from last 6 years. Finally G1 is a default collector in JDK9. From 2009, G1 has improved a lot and the bench-marking says its the best we can offer :). Eclipse gives a very beautiful view of G1 in working.

Java 9 introduces a major enhancement for modularity that helps define modules - The modular structure brings out a new paradigm shift in the way the programmer would program. From the perspective of JDT and its friends, this change means a relook at the existing concepts - starting from Java Model, through other core pillars of JDT. With the inputs from the EclipseConNA 2016, the JDT core is undergoing a change to incorporate these change both conceptually and at the core implementation level. In this talk, we will provide a brief introduction to the Java 9,  especially to those parts which affect JDT, we would also "redefine" some of the concepts in the Java Model, and then we will talk about the challenges faced and the solutions provided in JDT as a whole for Java 9.

JDK8 adoption was the best Java Adoption. And when we talk of Java 8, we come across functional programming and lambdas. I saw many talks going around on how to use Lambdas. Yes, it is still a nightmare for many developers to use :D. Lets first focus on why to use Lambdas. What are the performance benefits by using Lambdas.

In this talk, I want to show you the performance analysis of Lambdas using Java micro benchmarking. Eclipse has excellent support for JMH. In a quick, we can see a Lambda and a similar non-lambda code benchmarking.

Agenda will look like this :-

1. Lambda implementation and InvokeDynamic

2. Capturing and Non-capturing lambdas.

3. Performance Matrix for both the lambdas.

4. Eclipse Plugin for JMH - How to use it for benchmarking.

5. A real demo for Lambdas. (Can't do much but surely 2-3 demos as the warmup takes time)

6. When to use Lambdas and when not to use.