Posted by kathysierra
on June 27, 2005 at 11:12 PM PDT
Java may be mature, but it's far from sedate. In fact, it might just be getting warmed up...
The old optimism is still here in San Francisco. Sure we've heard it all before: the huge HUGE "opportunity" in mobile, the "renewed commitments" with other big players like IBM, and "the really interesting things", as Gosling calls them, that Sun's customers are doing with Java... but still, we can't help ourselves from suspending disbelief and getting swept away once again by the enthusiasm.
And why shouldn't we? In McNealy's session today, he pointed out that for the first three years of JavaOnes he was absolutely hammered by the press for "crazy talk"-- the outrageous overhyping of Java. But now, ten years later, if you go back and read what he said... it was dramatically underhyped! So why not indulge in a little optimism for Java? We tend to focus sometimes on the predictions for Java that didn't pan out, or that took much longer than expected... but so much more than most of us expected has happened for Java in these last ten years.
We imagined a world where Java was in devices, and while the whole toaster and Java Dog Collar thing didn't happen (thank goodness), Java is in billions of devices, with millions of dollars being made from downloadable Java-enabled content. We aren't wearing our Java rings (although I still have mine), but Java is the smarts inside a gazillion smart cards (and at least one Boeing airplane).
We imagined a world where Java was THE client-side web language, and while that didn't work out, Java ended up playing a much more substantial web role, powering the world's most heavy-duty enterprise server-side apps.
Of course we all wanted Java in gaming consoles, and that hasn't gone quite as we'd hoped, but now the Blu-Ray standard means Java will be inside virtually every next generation DVD player! Downloadable AV content, it seems, could be an opportunity even greater than supplying downloadable mobile device content. It could happen.
We imagined a world where JavaBeans (remember 1998?) were going to be The Thing, and everyone and their grandmother would be able to use a simple authoring tool to wire up an app (complete with multimedia) in no time. That didn't happen, and for a while it looked like the JavaBeans team went down to something like... 1/2 a person. But now 2005 is starting to look like 1998 again, only this time they really mean it. NetBeans is seriously rocking, and the client/desktop component model is showing up in all kinds of JSRs and tools. Sun's damn serious about adding those 5.5 million more developers, and they're pulling out quite a few stops to lure (or as Gosling put it last night, "seduce") them in with friendlier tools.
And color me naive, but I honestly believed it when Scott McNealy said -- and this is pretty close to an exact quote -- "We're absolutely committed to Jini, and supporting the community like crazy." He went on to explain that there's just so many things to talk about, and so much noise, that he and Sun haven't been doing a good job at talking about Jini (duh!), but that he wanted to do more. And he did, starting with saying, "Jini, Jini, Jini" into the mic.
He talked about how much Jini is used internally, and I suddenly remembered that four years ago (when I was a Sun employee) I wrote a Jini browser, and when I finally got it working I suddenly "discovered" all kinds of Jini services the storage guys were doing somewhere on campus. I was so thrilled to find real services to list in my toy browser that I never stopped to recognize that these services the storage folks were doing were serious Jini apps. (I, on the other hand, was building Really Important Applications like the Jini-enabled virtual pet kennel).
Anyway, all the optimism here at JavaOne doesn't mean we aren't a little more jaded, skeptical, and rational than we were in 1997, but when you look back over the last ten years, Java has changed the world. And y'all are a part of that.
(The highlight of the first day for me, though, was seeing the bullet point on the Mustang Features slide that mentioned wildcard classpaths. It's amazing how much applause one simple little asterisk can generate...)
So just pause for a moment and let it sink in how far we've come in ten years. Then try to project ten more years into the future. If the next ten years are anything like the last, then whatever you're imagining is surely an understatement.