I get random failures in DateTimeTest.testNow verify statement because of rounding, for instance:
-- Run: testSys::DateTimeTest.testNow...
dt1=2010-01-24T12:36:23.629Z UTC, time=12:36:23.63, dt2=2010-01-24T12:36:23.629Z UTC
I added echo statement before verify in the original sources the problem would be on line 173.
TEST FAILED
sys::TestErr: Test failed
fan.sys.Test.err (Test.java:179)
fan.sys.Test.fail (Test.java:173)
fan.sys.Test.verify (Test.java:57)
fan.sys.Test.verify (Test.java:54)
testSys::DateTimeTest.testNow (DateTimeTest.fan:174)
SideNote: a "coredump" of local variables in test method would be handy in case of a failure. I had to run the test in a long cycle after adding an echo to reproduce this issue.
brianMon 25 Jan 2010
I pushed a fix - let me know if you still see it after the next build.
katoxMon 25 Jan 2010
It helps thank you. Pretty tricky patch - I wouldn't guess that Time.now and DateTime.nowUtc do use caching but Time.now(timezone) always creates a new instance.
katox Sun 24 Jan 2010
I get random failures in
DateTimeTest.testNowverify statement because of rounding, for instance:I added
echostatement beforeverifyin the original sources the problem would be on line 173.SideNote: a "coredump" of local variables in test method would be handy in case of a failure. I had to run the test in a long cycle after adding an echo to reproduce this issue.
brian Mon 25 Jan 2010
I pushed a fix - let me know if you still see it after the next build.
katox Mon 25 Jan 2010
It helps thank you. Pretty tricky patch - I wouldn't guess that
Time.nowandDateTime.nowUtcdo use caching butTime.now(timezone) always creates a new instance.