Pixel complete

July 20th, 2010

In an indirect response to my earlier post, David Grace writes:

All this talk about not being able to create something that looks good in JavaFX is hot air. JavaFX has the functionality to do so, you just have to know how. What JavaFX needs is for the preview controls and layouts to be finished, controls such as a table implemented, and the Prism renderer implemented. When this is done it will be easy to write any application that could be written in Swing quicker, looking much better and with far greater performance. Personally i would of rathered investment in Swing instead, but Sun/Oracle have invested in JavaFX and in many ways its already far better than Swing, except for missing components.

Modern programming languages provide abstraction layers and higher level concepts to shield the programmers from the low level drudgery of writing assembly code. In much the same way modern UI toolkits provide component libraries, image composition, transformations, animations and many other tools that make it simpler to translate the intended design into a living and breathing reality. But eventually, it’s pixels. All the way down.

Just as the vast majority of the programming languages are Turing complete, the vast majority of UI toolkits are what i call pixel complete. If the specific toolkit exposes the ability to control every single pixel in the window (with, say, paintComponent in Swing, onDraw in Android, paintControl in SWT or updateDisplayList in Flex just to name a few), there are no limits to what can be put on the screen. Modulo support for translucent and shaped windows, you can take any design and turn it into reality. Indeed, if you know how to do it, you can do anything. But that’s not the point.

How about this – instead of asking what people can do with JavaFX, ask what people are actually doing with it.