Opinion: JavaOne 2011 Was a Success
After a difficult first year in charge of JavaOne in 2010, most people I spoke to at JavaOne this year felt Oracle was having more success with the conference this time round. The vendor really needed to show a broad spectrum of the Java community that it had successfully integrated Sun's Java teams, and was starting to drive the platform forward again, and it succeeded in those aims.
As well as the expected strong emphasis on cloud computing, there was a more unexpected, but very welcome, emphasis on both community and client-side Java. From the community stand-point, Oracle provided a briefing and brunch to JUG leaders and Java Champions on what was going to be in the keynote, as well as handing the final keynote over to them. “I...get the feeling Oracle has really energized the JUG leaders and the wider Java community,” wrote Eclipse Foundation Director of Marketing, Ian Skerrett, continuing,
The closing keynote session was all about community, including a session with the Java Posse. Good things are going to happen moving forward. Unlike when Sun was in control, I actually feel like the Eclipse community is welcomed into an Oracle-led Java community.
For client-side Java there was a strong emphasis on JavaFX. Stephen Chin, co-author of the book 'Pro JavaFX Platform', remarked
For those naysayers who said JavaFX is dead, Oracle has proved that they are going to continue to move the platform forward and support state-of-the art development on Java client technology. As always, Jasper and the rest of the JavaFX team had some pretty impressive demos and the entire Java community was re-energized about the future of client Java.
Attendance was also up according to Oracle. Although the vendor isn't able to provide separate figures for OpenWorld and JavaOne, the combined figure for this year was around 45,000 attendees, and JavaOne attendance was estimated at double that of last year.
There were certainly some weaker aspects. Google's continued refusal to allow its employees to speak at the conference meant, for instance, that former Sun/Oracle man Tor Norbye, now part of the Android team at Google, wasn't allowed to join the rest of the Java Posse on stage for the community keynote, a move which seemed rather petty on Google's part. Equally, Oracle's refusal to mention Android by name, even when demonstrating JavaFX running on an Android-based Tab (referred to only as a “Linux-based device”), looked faintly ridiculous.
There were other problems too. The keynotes, although certainly packed with plenty of information, felt a bit lacklustre, with way too many vendor-sponsored sections of little or no value for attendees.
In addition, holding the conference across three hotels meant there was no central place for people to congregate, reducing opportunities for both networking and just interesting casual conversations. Moreover, people staying in any of the Fisherman’s Wharf hotels had a lengthy shuttle bus commute of around an hour, with a change of bus at Moscone, to get to and from the venue.
Martijn Verburg, Community Leader for the London Java User Group, blogged that, "the speakers' room was very small and has limited WiFi signal". We had similar problems in the press room. More generally the WiFi, throughout the conference, was horribly unreliable, causing a great deal of frustration for attendees, making it difficult for presenters trying to do live demos, and for press, bloggers and tweeters alike to get stories out from the conference in a timely manner. To compound this, from the press point of view there was only one briefing event in four days, and only limited opportunities to talk to people from Oracle, again making it more difficult than we would have wanted to get full coverage of the conference out during the week.
I also honestly can't see the value in having combined JavaOne and OpenWorld. The two conferences have entirely different audiences, and a radically different feel. There is little or no sign of JavaOne attendees heading to OpenWorld or vice-versa (though perhaps if you held the two conferences closer together you would see more of this). JavaOne, at least in Sun's time, was always a profitable conference on its own, so why not simply hold two? That would allow Oracle to move JavaOne back to Moscone, where it surely belongs.
But, despite these gripes, the conference was a good incremental improvement on last year. Let's hope Oracle continues to improve the conference next year. For InfoQ's part, we'll be publishing more detailed articles on the three central themes from JavaOne, along with other interesting information from around the conference, in the coming weeks.
Java 8 delayed
Re: Java 8 delayed
Re: Java 8 delayed
Re: Java 8 delayed
Here is latest Java posse episode . You'll note that when asked if everyone agrees with pushing Java 8 out the reaction of the crowd is negative.
Here, is an interview with Mark Reinhold from the recent Javaone . I am petty sure after listening to it you will get the impression that it has less to do "with community" wanting a two year release cycle and more to do with not being ready for a 2012 release.
Furthermore, if Oracle cannot keep their commitment for a 2012 release what makes you so sure we'll have 2 year releases?
You should know that I think Oracle is doing great things for Java. It’s just disappointing to have to wait that much longer for Java 8.
Re: Java 8 delayed
"Furthermore, if Oracle cannot keep their commitment for a 2012 release what makes you so sure we'll have 2 year releases?"
Just to be clear I'm not convinced about that - I was merely pointing to Alex Miller's tweet. But I do think a regular rhythm of releases for Java would be welcome, and I suspect, and hope that Oracle will make the necessary resources available to do this. At the moment, there's still a lot of inherited problems from the end of Sun's tenure and the acquisition itself which won't be helping scheduling matters.
With regards pushing JDK 8 back 6 months it is certainly disappointing. But I will say that both the big ticket items in Java 8 - lambda and Jigsaw - are enormously problematic for different reasons (Lambda mainly because of the way generics was done, and Jigsaw because of having to maintain some sort of compatibility with OSGi). Given that my own inclination is to say "take as long as you need." Lets try and get these as good as they possibly can be. I appreciate this is probably the minority view.
Java for the Enterprise