> On Aug 5, 2015, at 5:06 PM, Jens Kapitza <j.kapi...@schwarze-allianz.de> > wrote: > > Am Dienstag, 4. August 2015, 13:26:06 schrieb Scott Palmer: >> The issue is: https://bugs.openjdk.java.net/browse/JDK-8090438 >> <https://bugs.openjdk.java.net/browse/JDK-8090438> >> >> A target release has not been set. But I really hope it could make it into >> 9.0. Has there been any further discussion on this? >> > Why not like this? > > https://github.com/sarxos/webcam-capture/tree/master/webcam-capture-examples/webcam-capture-javafx >
When I ran the JavaFX example on OS X, it locked up before even showing a UI. (It may have something to do with accessing Swing/AWT classes off the Event Dispatch thread, or before such thread exists*.) It’s also a pretty heavy library with way too many dependencies. It cobbles together or relies on 10 different APIs/libraries, or “drivers" to get the job done. In short it’s a mess compared to what the camera API should be. For reasonable camera access Java needs something much simpler. Cameras are built-in to most of the devices Java runs on, and have been for some time. Heck, even the Raspberry PI has a standard camera module and it runs JavaFX. There should be an official API for camera access. *On a side note, it’s a shame that there is no Image data type that is independent of JavaFX/Swing so manipulating image data doesn’t have to bring in a particular UI toolkit, and therefore force use of Swing even though your API is JavaFX based. E.g. ImageIO requires Swing-based images when it should have been independent of the UI toolkit. A Raster interface could be factored out and shared between whatever UI image classes are used, ImageIO, and the future camera API… probably too late now though. Regards, Scott
