A heap of interesting links this week, so I hope there is something of interest in the links below for you! As always, feel free to email me your links if you think they are worth sharing to everyone else. Let’s get into things!
- The JavaFX 2.0 beta builds keep on rolling off the production line. This week we have build 40 ready to be tested. From my point of view, the most exciting thing about this release is the performance improvements we’re starting to get into releases. Our specialised performance team are finding memory leaks and CPU hogs, and the engineers are doing their best to get things as performant as possible. In particular, b40 includes a heap of performance improvements to CSS and controls performance. Of course, there are plenty of bug fixes, and the general polishing that is expected of a relatively late-stage beta release. As always, my plea remains: please get it onto your machine and give it a good thrashing. Bug reports and complaints are highly appreciated!
- Dean Iverson has been busy posting two screencasts about building GroovyFX applications on to his blog. The first screencast is about how to get started with Griffon, GroovyFX, and JavaFX. The second screencast is concluding the basic introduction, introducing features such as binding, automatically generating properties, and keeping your JavaFX user interfaces as succinct as possible using the power of Groovy.
- We posted an updated ‘Introducing FXML‘ document. This should hopefully answer some of the questions you have.
- If the technical documentation above isn’t enough, there is also a ‘Getting started with FXML‘ article on the Oracle JavaFX website. This article works through the process of building a simple, example application and gives a good overview of how to work with FXML.
- And if that documentation isn’t enough, the JavaFX documentation team at Oracle have just posted a blog about all the new and updated documentation that recently got pushed online. They’re doing an excellent job making sure there is a lot of high-quality documentation about JavaFX 2.0, and my hats are off to them, knowing how much the engineers have increased their workload by changing APIs over the last few months.
- Tom Schindl has released e(fx)clipse 0.0.3, which includes CSS validation improvements for customising the look of your applications.
Have a great week everyone 🙂
I don’t really have the time currently to test the latest version out so I hope you can help with these questions.
Can you tell me if the video player can now play any video codecs that the OS supports? If not, is that coming up any time soon?
Is the Mac release going to be any time soon? What’s halting it to be released?
Can the webview view Adobe Flash or Silverlight?
I am waiting for this answers too.
Where are the menus in javaFX Docs ?
My bad, menus are under javafx.scene.control.Menu but no graphic examples in javaFX site.
When the Mac OS and Linux versions are coming?
we need the mac version
Thanks for the update, but i have some complaints about this release:
1. rendering is A LOT slower (up to 10x) than b34, especially when using drop shadow or blurring. The best example is to try ColorfulCircles demo, which basically doesn’t run well at all in this release.
2. Fonts are randomly reset to default style when mouse enters the scene.
3. Paths are really jaggy; anitialiasing is not doing enough.
btw, the mouse wheel api is now deprecated, the the demo codes are still suing them. Which new api should i use for mouse wheel event in this and future releases?
finally, i’d like to request some faster api to transform local coordinates to designated parent’s coordinate system. Currently you have to iteratively call localToParent which allocates a new temporary Point2D each time u use it. Considering the case when there’re 500 nodes’ coordinates need to be transformed, every of them is located in a deep scenegraph tree structure, this subsequently results in 5000 localToParent calls per frame just to obtain coordinates. You can feel how slow it is.
@truvel, regarding the coordinate transforms I hear you. We actually use mutable point objects internally for doing these transformations to avoid creating garbage, but the mutable objects aren’t part of the public API. Another option would be to have a “cumulativeTransform” property that is read-only and represents the entire transform from parent-to-local space, and then you can call methods on that, possibly. You can file an RFE though and we’ll add it to the queue.