<AWT Dev> xawt/libmawt.so and headless/libmawt.so

David Holmes David.Holmes at oracle.com
Wed Apr 20 05:33:33 PDT 2011

Alan Bateman said the following on 04/20/11 21:55:
> David Holmes wrote:
>> It's a different definition of "headless". The JDK can have headful 
>> capabilities but operate in headless-mode - all the libs are still 
>> present. The SE-Embedded headless JRE has had all the headful libs 
>> stripped out. We use the absence of the libs to recognize that this is 
>> a headless-JRE and use that to set the java.awt.headless property so 
>> that the Java libs act in headless-mode. So the VM drives this.
> I guess I'm wondering if the absence check could be done in the AWT code 
> rather than in the VM.

The only glitch I see with that is that the 
GraphicsEnvironment.getHeadlessProperty code is Java code while to check 
for the native lib we'd need native code.

It's doable I guess, but is it worthwhile? What problem are we trying to 


More information about the awt-dev mailing list