While obvious questionable code still exists (Norm still starts threads during construction), I've managed to knock over numerous visibility issues.

The tests now look very solid, I'm waiting now for a test cycle (one test repeated 100 times) to complete, so far I've seen no more failures.

[java] Running com/sun/jini/test/spec/javaspace/conformance/snapshot/SnapshotExpirationNotifyTest.td

bash-3.00$ grep "tests passed" test_results14.txt
    [java] # of tests passed    = 1
    [java] # of tests passed    = 1
    [java] # of tests passed    = 1
    [java] # of tests passed    = 1
    [java] # of tests passed    = 1
    [java] # of tests passed    = 1
    [java] # of tests passed    = 1
    [java] # of tests passed    = 1
    [java] # of tests passed    = 1
    [java] # of tests passed    = 1
    [java] # of tests passed    = 1
    [java] # of tests passed    = 1
    [java] # of tests passed    = 1
    [java] # of tests passed    = 1
    [java] # of tests passed    = 1
    [java] # of tests passed    = 1
    [java] # of tests passed    = 1
    [java] # of tests passed    = 1
    [java] # of tests passed    = 1
    [java] # of tests passed    = 1

Previously:

bash-3.00$ grep "tests passed" test_results11.txt
    [java] # of tests passed    = 0
    [java] # of tests passed    = 0
    [java] # of tests passed    = 1
    [java] # of tests passed    = 1
    [java] # of tests passed    = 1
    [java] # of tests passed    = 1
    [java] # of tests passed    = 1
    [java] # of tests passed    = 1
    [java] # of tests passed    = 1
    [java] # of tests passed    = 0
    [java] # of tests passed    = 1
    [java] # of tests passed    = 1
    [java] # of tests passed    = 1
    [java] # of tests passed    = 0
    [java] # of tests passed    = 0
    [java] # of tests passed    = 0
    [java] # of tests passed    = 1
    [java] # of tests passed    = 1
    [java] # of tests passed    = 0
    [java] # of tests passed    = 1

Reply via email to