Well, this morning I started writing on the blog quite early (8:55), an unusual thing since at that hour either I'm going to a customer or I'm still sleeping :-) but the fire alarm woke me up at 7:25 (and there was smoke smell too!), so at 7:30 I was ready to escape with my photographic gear in one hand and my laptop in the other. Fortunately, the alarm stopped buzzing soon.
Yesterday I gave my presentation about blueMarine. After ten minutes of almost-panic (there were problems with the video projector, but this time it wasn't Apple's fault) everything went smooth and in the end I didn't have to skip but a handful of slides toward the end. I think for sure it has been my best blueMarine presentation so far. I had 80/90+ attendees which was pretty good, considering that Gosling, Gafter and Bloch were talking about "the future of computing" in the other room. At last I can relax: there were indeed no worries about the presentation itself, rather about my health status (a few days ago I wouldn't have bet about my presence at JavaPolis). There are some other (technical) things I'd like to say about this latest demo, but I'm postponing them to another post.
To my surprise, almost half of people who attended my talk were directly interested, besides rich desktop applications, in Java and imaging! I think it has been the first time this happened in one year of presentations. So I unofficially announced imaging.dev.java.net, an initiative by Moritz Petersen, Emmanuele Sordini an me focused on the creation of a real community for Java imaging programmers. The idea is to build a shared discussion space for people using the many Java imaging technologies (Java2D, JAI, ImageJ, JImageMagick, JHLabs) since we feel today there's a lot of fragmentation - I mean, there are communities for the specific technologies, but not a common place to talk. imaging.dev.java.net has been just created, but some contents should appear soon. We will start with a list of all the existing commercial and open imaging applications made in Java, then we'll start interviewing their owner/developers about their experience - I hope they will be available for this initiative. Of course, imaging.dev.java.net is an open effort, so we're encouraging people to join and contribute. Don't let us alone.
The day ended with free french fries and free beers in the nearby pub, but it was a bad choice. I can't evaluate beers (I don't drink beer), but I hate crowded and noisy places where you have to scream to talk with people because of the loud "music". BTW, the previous evening we had the speakers/JUG leaders dinner and everything was perfect: quiet place, soft music in the background and you could enjoy chatting. Kudos to Stephan, Kristine and others.
I'm now listening to the Flex Keynote. I'm fairly cool about it. I see it as a backup for Java - at present time, if you need something working in a short time with effective, hot deploying and maybe requiring advanced media capabilities (e.g. playing movies) and - for future - should Java 6 Update N fail. To be fair, as an architect I could really advice about using Flex if you have pressing needs such as in the former case; but I don't see it as a strategic technology (I repeat, unless Java 6 Update N fails).
They're showing us a pretty demo of a "virtual" book that can be consulted by turning pages - pretty cool, but you can do it in Java too. Just look at the JBook demo from Bug's blog. Please note that pages in JBook can contain regular Swing components too.
Now, it's the turn of a Cover Flow viewer for photos - yawn, I've seen it three times yesterday: in JavaFX, in the demo by Angela Caicedo, in the Iris applet demonstration by Richard Bair and I've demonstrated it in a Rich Client Application. BTW, Java can do that by using JOGL, thus taking advantage of hardware acceleration, while I believe Flex can't (yet). Please correct me if I'm wrong.
So, everything boils down to the point I've mentioned earlier: easy, effective deployment (Flex is clearly a winner at present time).
Now, Flex has another clear marketing advantage: it's made by Adobe, which means it's used by people skilled in graphics and used to work with first class graphics tools such as the ones in the Adobe's suite. Thus, you'll see most of the pretty looking demos and products made with it. But if you're a Java developer, this doesn't mean you'll start making cool stuff out of Flex - and maybe you'll discover that you're not skilled in graphics at all (I just know a handful of people who master both programming and graphic design). This, anyway, is a critical point for Sun technologies. If Adobe is left alone in this area, in the end they will win. JavaFX must consolidate and then we need the graphic tools. I've repeated so many times, but the first move is not to have new, specific tools for JavaFX, rather plugins for the Adobe tools, so today's graphic designers can start using them. Otherwise, you'll end up having Java developers trying to play as graphic designers, and the disaster is assured. I've been repeating this many times, I hope Sun is listening, and if Sun isn't, maybe some third party can think about it.
For the rest of the Flex keynote, I got really bored and finished reviewing this blog instead of following it.
The remainder of the conference is due for the next blog post.
At last, today is the movie day, and they will show Beowulf. But, as usual, I'll be doing something else. Should I watch a movie, I prefer really one with some real photography, not a thing where virtual players are completely rendered at the computer (in that case, I'd rather watch some cartoons, assuming they are not boring clones of PlayStation games as most of the production of the latest years seems to be).
***** A necessary update. After the boring Flex part, Stephan kicked in. He demonstrated the new version of parleys.com (it's not yet online, now you can still see the old one). It's definitely a killer application - it's probably the best application around for enjoying video recordings of conference speeches with synchronized slides and advanced options, such as navigational and tagging features. It's made in Flex, of course, and at present time it has been the only choice around, as Stephen clearly explained. We're just falling in the case I've talked about: you need easy deploying, video streaming and your deadline is around now. One year from now we'll see if Java made it in filling the gap.