This Bugzilla instance is a read-only archive of historic NetBeans bug reports. To report a bug in NetBeans please follow the project's instructions for reporting issues.
Is anybody able to run recent IDE on recent IBM's JVM? http://www6.software.ibm.com/dl/dklx130/dklx130-p I got (an exception pattern): java.lang.NoClassDefFoundError: $Proxy1 sun.reflect.GeneratedConstructorAccessor9.newInstance(Unknown_Source) sun.reflect.DelegatingConstructorAccessorImpl.newInstance(DelegatingConstructorAccessorImpl.java:42) java.lang.reflect.Constructor.newInstance(Constructor.java:299) java.lang.reflect.Proxy.newProxyInstance(Proxy.java:585) org.openide.util.WeakListener$ProxyListener.<init>(WeakListener.java:963) org.openide.util.WeakListener.create(WeakListener.java:454) org.openide.loaders.FolderInstance$HoldInstance.<init>(FolderInstance.java:901) org.openide.loaders.FolderInstance.defaultProcessObjects(FolderInstance.java:644) org.openide.loaders.FolderInstance.access$100(FolderInstance.java:69) org.openide.loaders.FolderInstance$2.run(FolderInstance.java:592) org.openide.util.Task.run(Task.java:136) org.openide.util.RequestProcessor$Task.run(RequestProcessor.java:313) [catch] at org.openide.util.RequestProcessor$Processor.run(RequestProcessor.java:650) I wonder whether our bytecode (compiled using javac 1.4), IBM's implementation or my local setup is invalid.
Definitely bug in IBM's JVM. $Proxy1 and similarly named classes are generated by the JVM in the new (1.4) implementation of reflection. If the JVM claims such a class is not found, it can only blame itself.
Trung is right, not our fault - verifying, closing.
*** Issue 41301 has been marked as a duplicate of this issue. ***
*** Issue 35351 has been marked as a duplicate of this issue. ***
*** Issue 35080 has been marked as a duplicate of this issue. ***
Are we sure we are not doing something wrong? Apparently breaks OpenVMS too. Looks like some WeakListeners thing.
The only thing I am doing is that I am using the first, longer style for creation of the Proxy: http://java.sun.com/j2se/1.4.2/docs/api/java/lang/reflect/Proxy.html Instead of the second, shorter but more slower. I think that this is a bug of VM, however there is a workaround - in case this style fails, I am able to use the original, slower one.
Created attachment 14145 [details] Refinment of the talk to Proxy
Corrections first: This bug is not the same as the one reported by Meg on nbdev@. I have noticed that after writing my previous comment. This is failure in WeakListener class, not WeakListenerImpl one and moreover long time ago. It is not related to my changes in WeakListenerImpl between beta and rc1. I've attached a patch that may help Meg into this bug, as there is no better place to attach it.
Thanks Yarda. I'll see if this patch avoids the problem.
The patch works fine. Thanks again. Will this be committed to the release36 branch?
okay, we should try to get this in even though it looks more like a workaround for a bug in JDK impl. Otherwise OpenVMS users would not be able to use NB. Reopen and make this P2
Trung is right, this should be fixed to NB3.6.... Yarda, ask for review please, thanks in advance Thanks to Meg we've already verified fix, ...... Thanks Meg
My pleasure gentlemen. And thank you for your support.
Checking in WeakListenerImpl.java; /cvs/openide/src/org/openide/util/WeakListenerImpl.java,v <-- WeakListenerImpl.java new revision: 1.10; previous revision: 1.9 done
Backported to 3.6
*** Issue 41390 has been marked as a duplicate of this issue. ***
see Meg Garrison's comment -> verifying