Friday, December 14, 2007

JavaPolis - Conclusion

So what shall I say - it was an amazing experience being here in Antwerp @ the JavaPolis 2007. There were tons of interesting new technologies to discover and best of all meet other Java enthusiasts. I especially appreciated the open and friendly and personal discussions with (no special order) Neal Gafter, Joshua Bloch, Mark Reinhold, Chet Haase, Koshuke Kawaguchi, Brian Leonard, Stephen Coleburn, Guy Nirpaz and a lot more nice guys.

Special thanks go to Stephan Janssen - marvellous awesome work - hope to come back next year!

JavaPolis 2007 - Day 5

Already a bit worn out the last day @ JavaPolis 2007 began. Since the exhibition area was already closed down yesterday evening the crowd got together in even less space for breakfast, but that was ok.

First talk today was given by David Anderson about a kanban approach to software development. It is indeed a quite interesting approach but does not seem to work for all projects. Nevertheless you should check out his website and look for more informations about this.

Following this was a presentation about Java2D and 3D in high performance systems. Frank Suyken gave showed a couple of very amazing demonstrations based on JOGL and Java2D and gave away a couple of tricks how to make your own applications faster, ranging from GC to OpenGL.

After lunch there was a talk called "OSGi - The future of Java" given by Peter Kriens. He made some interesting points about actual problems concerning classloader and hiding implementations inside Java. On the other hand he seemed to insist that OSGi is the only true solution to this problem (ok he may be a bit biased). Nevertheless it was really interesting and will be stuff for a lot of discussions.

The last final presentation I attended was about "Unitils" from Flip Neven. Some quite innovative ideas how to make testing with databases more lightweight, but I will have to find out if this approach scales for large projects as well. But a few of the ideas are definitely worth considering.

It is a bit sad that JavaPolis 2007 is already over - would be nice if there could have been some gathering this late afternoon before anyone went of to catch their flights, trains or whatever.

Thursday, December 13, 2007

JavaPolis 2007 - Day 4

So on with day 4 of the JavaPolis. First thing is that you are starting to feel worn out. But this is ok - there is a price to pay for all this fun we are having. So even being maybe a bit tired we managed to get up early enough to get to the Metropolis in time for a good breakfast (seems I am starting to repeat myself, but it is quite astonishing even on day 4) and got into the keynote session - no need using the overflow room. Before the keynotes started the traditional JavaPolis Family Photo was taken - with all in the JavaPolis T-Shirt - but we have to wait for the result.

So the first keynote was about Flex. It was presented by Bruce Eckel with a lot of help from his friends from Adobe. It contained some definitive pretty slick UIs, but nothing that couldn't be done in java as well. The availability of the AMF remote protocol was announced, but there is nothing quite special about a binary remote protocol - I think we all have used one before. But you have to give them credit - the overall experience was quite amazing. Fitting in with this Stephan Janssen demoed the new version 2 - awesome UI (written in Flex), nice features and the best is you can use it offline and download all the Java talks to your local harddisk and play them from there. But describing the UI is not very useful - register for beta testing and see for yourself, if you are interested. The last keynote was done by Tim Cramer (VP Java FX Development at SUN) and was mainly focused on the Java FX product family and espceially the JavaFX Mobile Application Stack. It is a really interesting way to approach the cross device size problem - see for yourself by trying the different things out (all open).

The Java SE Update by Danny Coward had all the details on what is actually on the list of possible changes for Java SE 7 (and what is not - maybe). For details you can also visit his blog. Some interesting things seem to be coming and some language changes are also to expected (backwards compatible) but there is no decision what precise will be chosen. This directly takes me to the next talk from Josh Bloch - The Closures Controversy. Originally he planned to speak about Effective Java Reloaded - but he promised that he will directly start the work on the book when he gets home and will have it available printed and bound for Java One 2008. Instead of presenting about his new upcoming book he instead had a brand new talk with him about the benefits and risks of closures as actually proposed in BGGA specification (see yesterdays post).His main objections are that a adding closures and function types would add a new dimension of complexity to the Java language that are completely against the Feel of Java (as originally proposed by James Gosling in 1997). Indeed it is a quite scary syntax with a lot of fine tuning necessary how it should in deed work. Additionally there is additional overhead put to the developer until he can use this feature in a correct and concise way. So what is the point of introducing closures in such way? Just google it, go to Neal's blog and find out for yourself. So no conclusion on this - time will tell.

After this brain twisting presentation a quite fun event was up: The JavaPosse - Live Recording from the JavaPolis with Dick Wall, Carl Quinn

and Joe Nuxoll and Tor Norbye - the latter two were just available on-line ;-)

A lot of interesting things were discussed and at the end Neal Gafter and Josh Bloch were asked to join in for some questions and answers - that was real one of the entertainment hightlights of the conference - but find out for yourself at once it is out.

Next on the schedule was Stephen Coleburn presenting the JSR 310 - Date and Time API. Finally we are getting somewhere with the Date and Calendar problem - the API's look already real useful and quite compatible with JODA-Time, but you will have to change your code a bit.

So back to the big two rockstars Joshua Bloch and Neal Gafter. They did a nice presentation of the JavaPuzzlers - quite astonishing how many people missed often the correct solution, nevertheless real entertaining stuff. Get the book - I can only recommend it!

Then there was the prize draw at the SUN booth - but besides one Duke caught I did end up zip. Too sad - no UltraSparc as a continous buildserver for SQE. But hey that's life. So say good bye to our nice hosts atr the SUN booth (since it was the last day of the exhibition) and off the last BOFS of JavaPolis 2007.

First was the OpenJDK BOF presented by Mark Reinhold and Dalibor Topic. There were a couple of interesting things discussed from "How will all this with Mercurial work out" to Governenace, projects. It sounds like a real fun to be part there - sso stay tuned.

So getting already late in the evening the last BOF began: New Languages Features with - who would have expected this - Neal Gafter and Josh Bloch. They gave an overview about the possible additions to the Java Language and took a vote for each of the features presented. They hope that not everything they came up with, will necessary end up in the Java Language. Since the list was long it took quite some time to go through all those details. So it got really, really (45 min overtime) really late.

Because of this we once again ended up in the Hollywood Café drinking a good beer and discussing language changes with Neal and Josh. Thanks for sponsoring the event ;-) Can't wait to see any of those new spiffy neat (or whatever you might call them) features in Java. So that never again a kitten has to die because a Java Developer has to write Collections.<Man>emptySet().

Ah I forgot to mention tonight was Beowolf on for the conference participants - but hey isn't the JDK more sexy than Angeline Jolie - ok I have to think this over again ;-)

Wednesday, December 12, 2007

JavaPolis 2007 - Day 3

That was an interesting morning today. Everything went as planned and we arrived so early we had a lot of time left for a nice breakfast, filling out a survey at the SUN booth (getting a NetBeans 6 t-shirt that way) and answering some questions for the tomorrows edition of the Parleys magazin. Besides other very interesting things there are two things that grabbed our attentation - a NetBeans installation

and the friendly staff at the SUN booth providing free NetBeans 6 T-Shirts

Afterwards we got ourself a nice place to watch the todays keynote

- highlight: James Gosling.

There were some weird guys
and princesses out there

but in the end the keynote finally started. Before James Gosling took over we first had a very nice introduction from Stephan Janssen to JavaPolis and the future (Hope to come back next year). To make a long story short - these are the central messages I took with me from James Goslings' keynote
  • Java is Fun
  • Closures are sexy
  • NetBeans is better than sex
  • ..and something about when we last had an erection during presentation (maybe I got that wrong ;-) - perhaps I should replay the video)
Following this session we had a onverview about the things going on with OpenJDK - given from
Mark Reinhold. Pretty much is obviously in work - but lawyers aren't always helpful. But they are pretty sure to get all this encumbered stuff sorted out and a JDK7 project in OpenJDK has officially been started.

For lunch we had some hotdogs - so more lightweight food today.

The afternoon started with a further UI centric presentation Filthy Makeover by Chet Haase. It covered his Filthy Rich Client and Extreme GUI Makeover presentations and show some real slick things you can do with gradients and other Java2D effects. With this done a more project management session was on the agenda - Scrum in practice for non-believers. I hoped that this may turn me into a real believer - but it seems I am staying a skeptic for a bit longer since it does not seem to work for all projects. You have to have at least the right customer. The Future of Computing Panel was a fun thing to watch - the "Java Rock Stars" taking turns on picking at each other - felt quite familiar ;-)

Then there were two interesting BOFS concerning new language features for Java - Neal Gafter talking about the current state of his closure specification and implementation. It is fairly complete and can be download here. Give it a try and find out yourself if you love it or hate it. Afterwards Stephen Colebourne provided some infos on his new project which allows easy access to integrate new features into javac - for trying new things. You can find more info here. Afterwards we just called it a day and had some nice chatting over our free beer.

I forgot to tell you about the videos. All presentations are recorded and will be free available via So stay tuned - I think there was no time frame given.

Tuesday, December 11, 2007

JavaPolis 2007 - Day 2

Day two was the day of RIA - if you believed the session titles.

But before the sessions started we did a quick walk around the exhibition area. This contained a short stop at the Sun booth for collecting the appropriate t-shirt "May the Source be with you". A more thorough investigation is on the todo list for the next days.

We started with the Swinging RIA presentation (after having had breakfast), which included a nice demonstration of the power of the Beans Binding Framework and the Swing Application Framework provided from Richard Bair. Jeanette Winzenburg added some information on the current state of SwingX components and the challenges getting this to work in most natural way. Finally - for the first session - Chet Haase demonstrated the brand new Project Scene Graph (aka Scenario), which is a spin off from the JavaFX implementation. The project is currently in a really early early alpha state - so be warned, but the demonstrations shown were really impressive, but sadly it seems that they are still not available for download - maybe they may appear here some time later on (maybe we can find out more from Chet Haase tomorrow - or is it already today - time is running that fast if you are having fun).

After lunch there was a session about JavaFX - it was a lot of stuff that Jim Weaver put into his presentation. He tried to cover it all - from benefits using a declarative statically typed language to differences between the compiled and interpreted language.

So this was the bigger part of the day, but there was still more to come - a presentation about Ivy and short introduction on task oriented development (as supported by Mylyn) . The latter was a bit too Eclipse oriented - but this was to be expected from an Eclipse Evangelist.

Finally there was a quite interesting BOF hosted by Trolltech about QT. They gave some quite nice insights (you wouldn't believe why the name was chosen or the reason for using the prefix Q) on the QT history and has some demos showing the new Java API for QT (actually this seems to be some kind of automatic generated access layer).

Since tomorrow will be a very packed day - James Gosling, Mark Reinhold, Chet Haase, Neil Gafter, Bruce Eckel, Ola Bini - stay tuned for more info from JavaPolis 2007.

If you are missing some nice pics from todays sessions - there will be more pictures coming.

Monday, December 10, 2007

JavaPolis 2007 - Day 1

An interesting Day 1 of JavaPolis is nearing its end. So to get it all in sequence - first things first - breakfast. After having walked to the Metropolis from our flat this was a real good start into the day

After having had a real good breakfast (as you can imagine) we got to the conference rooms and I have to admit that the location is astonishing for such a conference. See for yourself

First session I attended today was The Zen of Agile Management.
This was quite an experience finding about all those things going wrong in software development and how to achieve a better process. After having lunch there was a real fun session with Brian Leonard and Oliver Nutter presenting JRuby and the Ruby on Rails integration in NetBeans 6.0, really impressive - and they were hurling t-shirts into the audience (the brand new NetBeans 6). Afterwards there was a session on code generation on a large scale using Freemarker and a presentation fomr Koshuke Kawaguchi on Hudson, a continous build integration server

For the short intervals between the session there was enough infrastructure to keep all the developers happy

The BOFs were very interesting as well. First a presentation about the OSGi approach (always a challenge for a NetBeans addict) and an astonishing presentation from the vice president of GigaSpace R&D about distributed systems, agile development and the special challenges. After a short discussion afterwards we called it day and went of for dinner, a Duvel and some more discussions about SCRUM, agility and the challenges of software development.

That was a lot for day 1 at the JavaPolis. More to come tomorrow.

Sunday, December 9, 2007

JavaPolis -The trip to Antwerp

Early sunday morning the trip to Antwerp started with a boat tour from Friedrichshafen with "The Katamaran"

to Konstanz

Then the train to Zürich

and via airplane to Brussels

From there once again by train to Antwerp

Finally we got to the Metropolis

There we already met some nice guys....

Stay tuned for more news from Javapolis

Friday, December 7, 2007

JavaPolis 2007

JavaPolis 2007 the first adventure - follow this blog closely to find out how to get there and for news on what is going at the Metropolis in Antwerp.

The journey starts on Sunday ;-)

Hope to meet you there.

New NetBeans Blog

So here we go again - a new Blog on NetBeans and Java. These are the new adventures of NetBeans 6 starship - going where no IDE has gone before.