> I was also wondering why this exception still passed the test, that's kind > of disturbing actually, and it's nice to know I'm not hallucinating.
I didn't go deep but it's probably not propagating up to the default thread's exception handler; it's caught, printed and discarded -- that's why the test passes. Normally an exception in a background thread will cause a test to fail. > like solr.core.name, solr.core.dataDir, solr.core.instanceDir and the > associated values. These were _always_ populated until SOLR-4196. Perhaps > Uwe's comment about other tests changing the environment is pertinent. Tests should be isolated, if they are not it's not good. We tried to make sure all (system) properties are restored back to what they were before the test is started. There are some known exceptions but in general system properties shouldn't cause order dependence among tests. Weird. D. --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
