I using Inline-Java-0.44 on a Win2k machine with Active State Perl v5.8.2.
The Java I am using is HotSpot 1.4.2-b28, mixed mode.
I am not using the JNI.

As a total Inline newbie I can try and say what I see but I can't explain a thing:

I see the same strange pause the Jim Hargrave reported for 0.43 at
10_1_shared_alone....ok 4/4
If I kill the java process then the tests continue with the other 'shared' tests being skipped.
Everything else passes (in fact '10_1_shared_alone' all passed too as shown).


I also get intermittent java termination errors from Dr. Watson after other tests when running the test suite.

If I run the example code for the Swing callback the example works but the program exit is pathological.
I do not get a Java error per se but a massive dump headed:


Full thread dump Java HotSpot(TM) Client VM (1.4.2-b28 mixed mode):
"Thread-0" daemon prio=7 tid=0x1833a7c8 nid=0x608 runnable [0..0]
"Java2D Disposer" daemon prio=10 tid=0x18300d48 nid=0x504 in Object.wait() [1a70
f000..1a70fd90]
...


This output including the addresses is consistent.
I am then returned to the command line but without the command prompt. Pressing enter recovers the command prompt.



It appears the callback loop is stable. If I comment out the call to StopCallbackLoop():
sub button_pressed {
print("click!\n") ; # prints click!
# $b->StopCallbackLoop() ;
}


Then I can click to my heart's content.

I hope this is helpful to some folks who might actually understand what is going on.

Alex Stoddard






Reply via email to