On 5/23/06, Andrew McIntyre (JIRA) <[email protected]> wrote:
[ http://issues.apache.org/jira/browse/DERBY-892?page=comments#action_12413064 ]
Andrew McIntyre commented on DERBY-892:
---------------------------------------
This looks good, but shouldn't we leave the permissions needed for running tests in useprocess=false mode in the base harness policy file, considering the recent work that was done with DERBY-1141? Since the permission is only granted on derbyTesting.jar, the tests are still quite secure when running in from the jars. You could have the harness append the necessary permissions if running in useprocess=false mode, but that might be more trouble than its worth. Also, this makes the lang/errorStream.policy file unnecessary, unless you choose to go the append-permissions-for-useprocess=false route.
Good thinking, thx.
Instead of creating a nist.policy file, I'll create a useprocessfalse.policy under util and append it when running with useprocess=false. I think that's cleaner than adding generic properties that will be used by all tests. I'll run derbyall and report.
Myrna
