On 2/8/12 12:44 AM, Knut Anders Hatlen wrote:
Rick Hillegas<[email protected]> writes:
On 2/7/12 8:38 AM, Knut Anders Hatlen wrote:
Rick Hillegas<[email protected]> writes:
My last checkin (1241479) re-enabled NativeAuthenticationServiceTest
on non-windows platforms. If the *nix tests fall over again tonight,
then we should disable NativeAuthenticationServiceTest again until we
understand the situation better.
Hi Rick,
I just gave the test a try with Java 5 on Solaris, and it seems to get
stuck in the fifth configuration (consistently). No problems with Java 6
or Java 7. Do you have any theories on why it behaves differently on
Java 5?
Thanks,
Hi Knut,
Thanks for running the test on other platforms. I don't know what is
special about that configuration. I have instrumented the test to
produce a fair amount of chatty output if you set
-Dderby.tests.debug=true. If you set that flag and run on Java 5 on
Solaris, we may get some clue about the problem from looking at the
last statements printed to the console before the test hangs.
I'll take a look. The latest test cycle (with the test re-enabled)
passed on all platforms with Java 6 (including Windows). With Java 5,
the tests only passed on Windows. No Java 5 results were reported from
the Solaris and Linux variants.
Does the test pass with Java 5 in your (OS/X) environment?
Can't test that. Apple doesn't want users to linger on old VMs. You
can't get Java 5 for my machine anymore. In its infinite wisdom, Apple
only put Java 6 on my old machine (when they repaired its crashed disk).
On my new machine I have Apple's official Java 6 and various versions of
the OpenJDK Java 7.
Thanks,
-Rick
