I have been traveling quite a lot in the last few months, specifically to help with the Java 7 launch. During very important events, like the International Free Software Forum, FISL, and the very successful The Developer’s Conference, TDC, where SouJava also hosted the Brazilian participation of the pre-launch of Java 7, I and many other JUG leaders, gave talks about Java 7, promoted the good will of the community around Oracle’s efforts, dispelled doubts that Java is losing its importance and promoted the openness represented by OpenJDK.
The party has been fun, with the participation of many Java User Groups, hundreds worldwide. In Brazil, Mauricio Leal, Yara Senger and others from SouJava worked with JUG leaders to help them get giveaways and speakers and to organize more then 20 events. I traveled to several places, side by side with Java’s amazing, (nearly) untiring globetrotter Roger Brinkley, and we worked hard to reached as many developers as possible. In Brazil we had Arun Gupta (that also visited many cities), Simon Ritter, Dalibor Topic, Sharat Chander and others. Many traveled around the world.
This is an important release, Java is really moving forward! So important that despise the fact that several in SouJava expressed concerns about issues in the JCP, we have collectively decided to put those aside and vote for Java 7. We are all putting our credibility in this, to push the technology forward.
I then traveled to OSCONJava, in Portland, and with Fabiane Nardon, talked about the Future Java Developer, playing along with the jokes of “the future Java developer is a retired developer!”. We don’t think so, Java is an exciting technology and has a strong, vibrant future. And Java 7 was really release during OSCON! Very well received into the community, with many “congratulations” messages from JUG Leades, Java Champions and from the NetBeans Dream Team. Lots of twitter posts, and praises from all around! We were all somehow part of this huge milestone for the Java Community! This is the first release ever to be based on an open source project, and OpenJDK is now the Reference Implementation!
But… what is this now? With all those travels (and I’ve visited Seattle and now Pittsburgh), I pretty much missed the next few days… and surprise… Java 7 has a bug, that corrupts data? Looks like it is because it uses by default some new HotSpot optimization? Ok, there is a workaround, I’m sure the workaround has been set by default in all new downloads? No? There is a warning? No? Georges Saab has sent an email to the Java Champions, this is good. But I could not find an Oracle statement for the general public. I did a quick (totally non-scientific) poll with 8 developers that were online. All of them knew there was a bug in Java 7, none of them knew what Oracle said about it. One of them searched and sent me a link that describes the problem. This is not good…
There are timid public mentions: Henrik quotes (in the comments!) Mark Reinhold’s info that it will be fixed (try!) in update 2. Another timid Oracle response came in the OpenJDK mailing list, where Vladimir Kozlov says “we’ll try… can’t promise…”. Outside from Oracle, both Cay Horstmann and @myfear wrote wisely that this is a minor issue. And of course, Uwe, from Lucene, is quoted saying “Oracle took action by changing the priority of the related bugs”. If you go to the known issues on JDK 7 you can see it there, but it does not really tell you that this can corrupt data, and please don’t run without the workaround! Not very hopeful posts…
And I agree with Stephen Colebourne (whose message was the first I saw alerting to this fact): no one seems to be talking about it… Are we afraid to damage this important release? This does not look like a huge a technical issue: the bug has already been identified, an workaround has been produced. But it affects at least one prominent and widely used project. This is political, and the response we see from Oracle is what will define the fate of this. It is Oracle’s response that we’ll be able to quote and replicate. It is the decisive, immediate, steps that Oracle takes to fix this, that will stop the rumors from spreading. And so far, from what I can see, the response is minimal… The news that there is a flaw in Java 7 that corrupts data will hunt us for many years. Unless it can be clarified now. And not soon, immediately. Before the blogosphere spreads this even more!
Not coming forward damages the credibility of a technology that powers the world’s largest enterprises. It damages the credibility of ever JUG leader, Java Champion, Java developer, JCP member, NetBeans Dreamer, that have spent the last couple of years fueling the excitement for the new release. It damages the Java team at Oracle, and the OpenJDK project. Please Oracle, do it, and say it, loud and clear, whatever needs to be done and said. And we can all go back to the party. There are still several “post-launch” events that are happening around the globe, lets make sure developers on those events hear the right message — and that they can download a Java 7 that they can trust.