On 24/May/2010 11:37, Mark Hindess wrote:
> In message <4bfa5296.4020...@gmail.com>, Tim Ellison writes:
>> This doesn't compile successfully for me using
>> Windows XP, MS VS2003, Ant 1.7.0
>
> I'm guessing you might have ant properties setting hy.zip.api=true which
> is not correct for DRLVM.
D'oh. Yes, I had set up some 'global' properties and forgotten. Sorry
for the distraction.
I'm able to compile ok, and the tests run through with a few errors.
Swing:
BoxLayoutTest#testLayoutContainerVertical
Test interrupted due timeout
BasicScrollBarUITest#testGetThumbBounds
(BasicScrollBarUITest.java:267)
Expected: 16 actual:0
NIO:
SocketChannelTest#testSocket_Block_BasicStatusAfterConnect
(SocketChannelTest.java:495)
expected not same
SocketChannelTest#testSocket_NonBlock_BasicStatusAfterConnect
(SocketChannelTest.java:495)
expected not same
JDWP:
OnthrowDebuggerLaunchTest#...
A number of these are failing with
org.apache.harmony.jpda.tests.framework.TestErrorException:
[SYNC] Exception in accepting socket sync connection
The console output for a number of the JDWP tests shows
[junit] [SYNC] Accepting socket connection
[junit] STDERR> Uncaught exception in main:
[junit] STDERR> java.lang.NoClassDefFoundError:
Files\Java\jre6\lib\ext\QTJava/zip;C:\Program
[junit] STDERR> Caused by: java.lang.ClassNotFoundException:
Files\Java\jre6\lib\ext\QTJava.zip;C:\Program
[junit] STDERR> at
java.net.URLClassLoader.findClass(URLClassLoader.java:907)
[junit] STDERR> at java.lang.ClassLoader.loadClass(ClassLoader.java:488)
[junit] STDERR> at
java.lang.ClassLoader$SystemClassLoader.loadClass(ClassLoader.java:870)
[junit] STDERR> at java.lang.ClassLoader.loadClass(ClassLoader.java:267)
[junit] STDERR> FAILED to invoke JVM.
[junit] Redirector completed: STDOUT
[junit] Redirector completed: STDERR
It appears that the classpath is not being quoted, and the spaces in the
path to my hosting JDK are being interpreted as argument separators.
I'm guessing that's why a number of JDWP tests are failing for me.
>> I'd also point out that the ZIP file sig is a 1024/DSA, and we are
>> being encouraged to move up to 4096/RSA.
>
> I'll try to fix this for the next release (and the arguably more
> significant problem that no one actually trusts me ;-).
Hey, I trust you. :-)
Regards,
Tim
