About three months ago i have posted an interview with Amy Fowler, one of Sun’s senior engineers working on JavaFX. In that interview i had a question that bothered me ever since the original announcement on JavaFX at last year’s JavaOne – what is its target audience? To my surprise, that question had by far the longest answer in the interview, and i still don’t have a clear answer.

Apart from an understandable wish from a for-profit company trying to break into a lucrative market, it is not immediately clear what is the problem that JavaFX is trying to solve. Before going on a slightly off-topic tangent of the well-designed applications in general, Amy says:

In terms of the general target audience, we obviously have to focus initially on meeting the needs of a specific community, which is the RIA crowd (is anyone else tired of acronyms?). They are clearly less resistant to trying something new, especially when it speeds up development and is built on Java, which has always been about the internet.

Implying that RIA developers are waiting en-masse to try JavaFX as soon as it hits the web just because it is built on Java doesn’t really sound realistic. Outside the “walled garden” of internal development, Java has little weight of excitement as far as the web-based client-facing technologies go. As the recent Sun alumni John O’Conner is finding out, business shops already have an established, proven and ubiquituous alternative solution – Flex and Flash:

I’m learning that web application software teams love Java, but they tend to relegate it to the backend, business logic space. Rich client apps? No way. How about Java Web Start? No need. So, how about just a little, teeny, tiny applet then? Get outta here. Flash, flex, swf files…sure, but no Java on the front end. This particular team won’t even consider Java there. I can’t even begin to argue against Flash at this point.

This brings me back to the topic at hand – what is JavaFX trying to solve? Is it the “boring” internal applications? These are highly unlikely to be rewritten, even if they are written in Swing. The “sexy” flashy browser-embedded content? Adobe has hordes of highly trained designers and developers backed by the extensive tooling and continuous improvements to the platform (including hardware accelerated shaders in Flash 10).

Modularized kernel, applets that can be dragged to the desktop, embedding code samples in the docs, binding and animations built into the language – these are just a few features of JavaFX and 6u10. The word “innovation” pops up in quite a few blogs and articles on JavaFX, and it is a very dangerous word that can easily distract from the end goal – addressing the real world problems. As Scott Berkun writes,

One telling anecdote in my research into innovation history is this striking observation: inventors, creators, and leaders, the people who earned fame for the work other people call innovative rarely used that word themselves. Instead their vocabularies leaned heavily on words like problem, experiment, solve, exploration, change, risk and prototype. Powerful words. Words that are either verbs, or imply a set of actions. And more to the point, they care less about being innovative than they do about making things. Making good things. Forget creating a breakthrough: it’s hard enough to make a really good thing that people will love to use. Most markets are in desperate need of affordable, high quality goods that live up to half the promises their advertising make for them.

The first SDK drop should be out by the end of this month. I hope that “by invitation only” from the (still) horrible javafx.com is just a cheap marketing trick, and we will be able to download it freely without registration of any kind. I hope that we are going to see real applications from real companies. One of these might come from MaldenLabs, as hinted by its CTO in the same exact comment here and here:

We are developing a very large scale enterprise application that will utilize JavaFX. Obviously we considered Flash/Flex and Silverlight, but in the end “bet the farm” and chose JavaFX. We have been pleasantly surprised with the advancement of the language and the unbelievable support and commitment we’ve received from Sun. All I can say is that people are going to be ~very~ surprised.

I also hope that these words would be backed up by replacing the Flash content on that site with a JavaFX-based applet. And finally, I hope that the end product will justify all the time not spent working on Swing.

In the grand scheme of things, it all doesn’t matter. Technologies die, new technologies are born, people move to other companies, old prejudices refuse to die and some decisions are forced on technical people. The customers, of course, couldn’t care less about all this.

Here are some Swing links that you might have missed during this week:

• Clemens Eisserer continues working on the XRender pipeline for Java2D, and has a few screenshots showing the progress of this project. Part 1, part 2 and part 3 show different test applications under this pipeline.
• A message from Richard Bair on the mailing list of AppFramework (reference implementation of JSR 296) announces that Alex Potochkin has taken over the ownership of this project. The time indeed has arrived – the last source commit was in October 2007, and while Alex hasn’t posted any message on the mailing list / his blog, i would assume that he is taking his time making himself familiar with the project itself and the JCP / JSR process in general.
• Alex himself continues writing about the JXLayer project and talks about the internal painting implementation in his latest entry. Jing Ge has a follow up entry in which he builds a complete validation module using BeansBinding, JXLayer and Hibernate Validator.
• Xavier Young has a short introduction for using the EaSynth look-and-feel.
• Ethan Nicholas introduces the JavaCSS project that aims to provide CSS support for styling Swing and JavaFX components. While the amount of comments on this entry indicates a significant level of interest in this project, it would be interesting to see how well will it be able to cope with the expectations implied by using the “CSS” in its name. As mentioned by Ben Galbraith in his JavaOne 2006 session, Swing and HTML are sufficiently different in that you can not take the existing “pure” CSS directives and apply them on Swing components. In addition, implementing the layout constructs of CSS may require changing the existing application layout managers. Throw the look-and-feel layer into the mix (don’t forget that the look-and-feel is responsible for computing the inner metrics of Swing components, such as margings, padding, gaps, etc), and you get quite an interesting challenge.
• Ken Orr writes about unified toolbar buttons in his quest to emulate the appearance of native Mac applications. Surprisingly, this entry features a rather questionable advice on using the SwingUtilities2 class. Not sure about the Apple VM, but this is a strict no-no for even Sun VMs. Not only this class is in an internal sun package, it has also changed location between JDK 5.0 and JDK 6.0. What’s the alternative for SwingUtilities2.drawStringUnderlineCharAt? Call BasicGraphicsUtils.drawStringUnderlineCharAt and install the desktop rendering hints before that.
• This is most probably will be a part of the next weekly JDK 7 build. Bug 6725214 on forward-porting the new D3D rendering pipeline from 6u10 has been marked as Fix delivered. For the brave souls wishing to venture into the implementation details, follow this Mercurial changeset. No word yet on porting the support for shaped and translucent windows and which package that might end up in.
• Jacek Furmankiewicz continues working on the Java SwingBuilder project, and this time he writes about adding an SWT-style API for creating menus.
• And finally, Richard Kennard adds component wiring based on Commons JEXL to his MetaWidget project, aiming at those environments that don’t have expression languages built into the language (such as Swing).

< Prev     Next >

About half a year ago Christopher Deckers had a guest spot post on this blog about a non-daemon WebStart thread that prevents JVM from exiting when you’re closing your last Swing window running under the DISPOSE_ON_CLOSE mode. To illustrate the problem with the current implementation of WebStart helper threads, he posted the following excerpt from the AWT threading documentation:

Prior to 1.4, the helper threads were never terminated. Starting with 1.4, the behavior has changed as a result of the fix for 4030718. With the current implementation, AWT terminates all its helper threads allowing the application to exit cleanly when the following three conditions are true:

• There are no displayable AWT or Swing components.
• There are no native events in the native event queue.
• There are no AWT events in java EventQueues.

There is an additional clarification to the last item in the same document:

Make sure that no method of AWT event listeners registered by the application with any AWT or Swing component can run into an infinite loop or hang indefinitely. For example, an AWT listener method triggered by some AWT event can post a new AWT event of the same type to the EventQueue. The argument is that methods of AWT event listeners are typically executed on helper threads.

So, as long as you don’t post another event from your event handler or hang indefinitely, you should be OK, right? That is not entirely correct, since the current implementation of the last item (at least in Sun’s VM) makes an additional restriction on shutting down the AWT queues in absence of events.

The relevant class in Sun’s VM is sun.awt.AWTAutoShutdown and it has additional logic on top of the official documentation (relevant part is highlighted by me):

The internal AWTAutoShutdown logic secures that the single non-daemon thread is running when AWT is not in ready-to-shutdown state. This blocker thread is to prevent AWT from exiting since the toolkit thread is now daemon and all the event dispatch threads are started only when needed. Once it is detected that AWT is in ready-to-shutdown state this blocker thread waits for a certain timeout and if AWT state doesn’t change during timeout this blocker thread terminates all the event dispatch threads and exits.

In the current implementation this timeout is 1000 ms (or one second). What this effectively means that AWT is not shutdown immediately after disposing the last window in your application and processing all pending events. Instead, it wakes every second, checks for any pending or processed events during the sleep and continues sleeping if there have been any such events.

How can this affect your application? In my case, i have a daemon thread that wakes up every 100ms and synchronizes the visual state of certain components. Since everything UI-related must run on the Event Dispatch Thread, i wrap each iteration loop inside the SwingUtilities.invokeLater (otherwise i run into sporadic deadlocks). As far as AWT is concerned, such an invocation is an event which effectively prevents AWT from shutting down.

Bottom line is that you should be careful with calling SwingUtilities.invokeLater from daemon threads in applications using DISPOSE_ON_CLOSE. Before calling that method you can check that the relevant component is visible and showing. If both are true it means that the window of that component is still not disposed.