Monday: 11:00 AM-12:00 PM: The Future of Java Build and Continuous Integration
Panel discussion - no meaningless slides - which was nice. Did not go much into technical details but was interesting at a high level. Discussion around validating commits reminded me how crude the state of the art is when it comes to reviewing, testing, and bisecting changesets; DAG-based SCMs just do not do this well.
Monday 12:30 PM-1:30 PM: Continuous Integration with Jenkins
Some graphs claiming to show Jenkins is outpacing Hudson in activity by an order of magnitude. Then on to a technical topic - validating commits. Nice discussion of possible approaches and their problems, including the options supported directly in the Git plugin, and some plans (?) for similar functionality in SVN. During Q&A someone asked for similar functionality in the Hg plugin.
Monday 4:00 PM-5:00 PM: Why Doesn't Java Have Instant Turnaround?
Largely a product pitch for JavaRebel. Not much technical depth but did a good job of explaining the problem to be solved.
Monday 5:30 PM-6:30 PM: Project Jigsaw: Find the Corner Pieces First
Mainly introductory Jigsaw usage; nothing much new if you follow the lists.
Monday 7:00 PM-7:45 PM: Using Griffon to Build Groovy-Based, GUI-Based Desktop Applications Like a Wizard
Seemed to assume a Grails background, and font size was too small to be legible until I get a new eye exam.
Monday 8:00 PM-8:45 PM: What Functional Purists Don't Want You to Know About Scala
Just slides, but a decent summary of useful things in the Scala language that need not intimidate people without a theory background.
Monday 9:00 PM-9:45 PM: (nothing, too tired and everything looked boring)
Tuesday 12:00 PM-1:00 PM: Le Tour de xUnit
Sounded promising, but speaker used the phrase "pretty much" several times per sentence and I could not pay attention. Code also illegibly small. Resolved to get seats near the front for subsequent sessions, though this is hard if you are running late.
Tuesday 1:30 PM-2:30 PM
(nothing, prepping to talk)
Tuesday 3:00 PM-4:00 PM: Next-Generation UI: JavaFX 2.0 and Scala?!
Some interesting content. Would have been better if it had focused on showing plausible FX apps and how they can be written concisely in Scala, rather than wasting time on Scala language arcanalia and discussion of internals.
Tuesday 4:30 PM-5:15 PM: The Right Decisions at the Right Time with Jackpot 3, Codeviation, and Archie
Codeviation & Archie never materialized. Interesting discussion of Jackpot 3, though lists of syntax elements would have been better replaced by showing examples of usage. Some demos but not enough. Audience naturally asked about Eclipse integration and CI support, neither of which are available (yet?).
Tuesday 5:30 PM-6:15 PM
(nothing, prepping to talk)
Tuesday 6:30 PM-7:15 PM: Mavenize Your NetBeans Rich Client Platform Application
(That's me!) Went smoothly but quickly ran out of time and only got to show maybe a third of what I wanted. Attendance not large but mostly retained. Questions included third-party library wrapping (demoed), and NBI usage from command line (not well supported yet).
Tuesday 7:30 PM-8:15 PM: Monitoring a Large-Scale Infrastructure with Clojure
Interesting talk with good Q&A. DevOps automation using Clojure fragments for filtering and aggregating statistics. Would have liked to see more code since what was there seemed concise and expressive.
Tuesday 8:30 PM-9:15 PM: Using Java and Genetic Algorithms to Beat the Market
Sessions about trading can be iffy. Lots of discussions of financial strategies and the market which were irrelevant for this conference - I do not care whether you made or lost money, I am here to learn about Java. But, some code shown and demoed, and discussion of GA using Watchmaker was interesting.
Wednesday 8:30 AM-9:30 AM: Building Secure, Interactive, Scalable Web Applications Easily on JVM with Lift
Organized, plenty of demo time, good Q&A - worthwhile session.
Wednesday 10:00 AM-11:00 AM: The Enterprise RCP: NetBeans RCP with JavaFX 2.0 Controls
Potentially interesting topic but presentation focused on the implementation code of the integration, not relevant to the audience.
Wednesday 11:30 AM-12:30 PM: NetBeans Project Management
Introductory level but audience seemed engaged. Presenter discussed freeform projects; I brought up automatic projects during Q&A.
Wednesday 1:00 PM-2:00 PM: From DevOps to NoOps: 10 Cloud Services You Should Be Using
First part of session, overview of innovative services, was interesting and fun. Latter part, promotion for speaker's own product, was not very compelling.
Wednesday 3:00 PM-4:00 PM: OpenJDK Development Best Practices
Thursday 11:00 AM-12:00 PM: Calling All Modularity Solutions: A Comparative Study from eBay
Interesting roundup of module systems (NB not included). Bizarrely included "Maven" as one choice, though this is of course just a build environment and complementary to runtime module systems. OSGi (predictably) came out on top but speakers lamented the fragmented tooling, and also the difficulty in reusing pre-OSGi libraries using class loader hacks.
Thursday 12:30 PM-1:30 PM: Get Efficient in Building and Diagnosing Modular Applications with Eclipse Virgo and OSGi
Discussion of "snaps" occupying first portion of talk was of no interest. (Seems a very crude substitute for general services.) Hoped to see information on how to modularize enterprise apps and diagnose class loading issues, but there was very little here; a waste of time.
Thursday 2:00 PM-3:00 PM: LMAX Disruptor: High-Performance Concurrent Programming Framework
Zero code shown but very interesting nonetheless. Description of how to get sub-millisecond latency in a specialized Java server even on a single (pinned) thread by paying careful attention to hierarchical memory cache usage, avoiding concurrency primitives whenever possible, reclaiming heap storage, and similar techniques. Anyone who claims that "real programmers do not use Java" should pay attention.
Thursday 3:30 PM-4:30 PM: Who Said Runtime Class Reloading Is Hard in Java?
Javeleon. Nice demos though presentation was too fast and left much necessary background information unstated. Some interesting Q&A, much of it from attendees not grasping that there is no server support yet.