Swing links of the week: December 21, 2008
Here are some Swing links that you might have missed during the last week:
- Heidi Rakels has announced release 1.4.0 of Sanaware Java Docking library. Notable new in this version are support for applets and bug fixes for GTK and Nimbus look-and-feels.
- Jennie Hall has an example on using progress monitoring APIs with SwingWorkers. Read comments by Richard Osbaldeston on EDT-related violations.
- Geertjan Wielenga picks on the news of improved heavyweight / lightweight support in 6u12 b02. In the comments Gili Tzabari points bugs 6776743 and 6783411 that are still open in the area. Also, as confirmed on this forum thread, this support introduces a performance regression of around 1.5% even for applications that do not use heavyweight components.
- Jeanette Winzenburg has some strong words regarding the “JavaFX == Swing 2” in the comments on Chris Adamson‘s editorial blog on java.net, with intense followup discussion on omitting BeansBinding from Java 7 continuing in this java.net forum thread. If you are interested in this subject, an “embedded” discussion between Osvaldo Doederlein and Richard Osbaldeston in the comments on Java 7 Update is worth reading.
- Remy Rakic has been hard at work trying to decipher the JavaFX internals and allow embedding JavaFX content in Swing top-level windows. The code is linked here, and you can see a sample screenshot of embedded content here.
- Clemens Eisserer continues his work on XRender pipeline (scheduled to be part of Java 7 – congratulations), reporting that most of the native C code has been ported to pure Java2D.
- Roman Kennke is another winner of OpenJDK Innovators’ Challenge, and he continues his work as well. He provides a UML architecture overview of Caciocavallo project, and talks about enhancements to managed windows.
- Jeremy Wood tackles the problem of exporting layered vector-based graphics to a serializable format.
Finally, Mikael Grev has announced release 3.6.2 of MigLayout manager for Swing and SWT. Mikael has been vigorously pushing for inclusion of this project in the core JDK distribution, amassing an impressive number of 289 votes in the bug parade. I am not quite convinced that it will be a good match. Looking at the available release notes history (since June 08), i see quite a few bugs being fixed (even this very announcement was made on version 3.6.1, with a bug creeping in requiring another minor release), as well as improving support for IDE integration. This is a perfectly normal situation for any project, even when the APIs are frozen as far as the project developers are concerned. Would you want to forego quick bug fixing cycle for a larger audience? Would you want to risk the restrictions of JDK (never breaking the APIs) even you are positively sure right now that this is the final state? These and other related questions don’t have a definitive answer (to witness quite a few API disasters in the core JDK), and tomorrow’s interview (not related to MigLayout) will provide an interesting take.