With the 2007 JavaOne Conference four days away, I thought I'd share some session recommendations with my fellow attendees. Since Arun Gupta beat me to it, you should take a good look at his choice of sesssions, which has a large overlap with mine, then take note of a few differences. I bet you don't need me or Arun to tell you to attend Ben Galbraith's Ajax talk (TS-6836) or that other perennial favorite, Josh Bloch's "Effective Java Reloaded" (TS-2689).
This year I've served as the track lead for the Java EE track at JavaOne, so that's where I'll be spending most of my time at the conference. Incidentally, I'd like to thank all the submitters for flooding us with their papers and the review team (including Hani Suleiman as an external reviewer) for their work over the past few months. This is not the place to share tips on how to get your session accepted at the conference, but I can't resist dropping a few ones. The review team is very diverse, we've all attended many conferences over the years and we think we can spot a bad submission at a safe distance. Good submissions are interesting, fresh, well-written, with detailed and convincing outlines, and they include supporting materials in the form of web sites or blogs. We routinely check out the online documentation of any technology we are not familiar with. If you are talking about something new, make it compelling; if it's about a technology that's been around for a while, be thorough and go into some real depth. The abstract needs to communicate that your talk is of consequence to developers: JavaOne is not OOPSLA. Above all, be interesting: given the number of submissions, if nobody on the review team feels strongly that your talk is worth having, ultimately it will be rejected.
In the past few days I've been monitoring the registration numbers and it's clear that the Java Persistence API and Enterprise JavaBeans 3 talks are the main draw, as we expected, so you'll need to register ASAP using the online schedule builder. Two other talks are almost completely filled already: TS-4249 ("The Top 10 Ways to Botch Enterprise Java Technology-Based Application Scalability and Reliability") and TS-1262 ("Killer Apps: Data Mining Demystified"). If you are interested in them, hurry up.
This year we had a smaller allocation of BOFs than in the past, so we created a 2-hour super-BOF for all Java EE technologies. It's listed as two separate events: "Java Platform, Enterprise Edition (Java EE): Meet the Experts (Part 1)" (BOF-4641) and Part 2 (BOF-4612). Although I'm listed as a speaker for both, thankfully you won't have to sit through two hours of elocutions by yours truly; instead, I'll be playing host to a full lineup of specification leads for the different technologies in EE. The plan is to give a short intro to the main components in the platform and then take questions from the audience. So if you want to discuss anything EE-related, from the high-level roadmap to the most minute details of some technology, sign up now, show up at the BOF on Wednesday night at 7:55pm and grab the mike.
As for the other tracks, anyone interested in Phobos should come to the session that Ludovic Champenois and I will share on Thursday at 4:10pm (TS-6957). Since we'll be demoing but not fully describing jMaki, it'd be helpful if you could go to some of the jMaki sessions scheduled well in advance of our talk, e.g. TS-6375 and TS-9516. If you can stay up late (kidding...), there is a BOF on Grizzly on Tuesday at 10pm (BOF-6012) that sheds some light on how we embedded Grizzly in NetBeans to deliver the small-footprint, fast-startup Phobos development-time container. Also, TS-2992 describes in gruesome detail the entrails of the beast (I may have gotten carried away by Jean-Francois Arcand's bear imagery...); anyway, highly recommended to all NIO propeller heads.