On Jan 26, 2007, at 2:31 AM, Naveen Neelakantam wrote:
On Jan 25, 2007, at 4:13 PM, Geir Magnusson Jr. wrote:
Ok - modified.
Now CC is running w/o failures in standard SVN configuration on
Ubuntu 6 running on em64t
Naveen, you should have no problems either...
You were correct sir. The classlib tests passed for the first time!
Should I try getting rid of the perTest flag?
Try it. See what happens. I'm running ok w/o it on unbuntu 6 on an
older dual proc box.
Also shouldn't emails from my CC have shown up on the alerts list
by now?
I was just pondering the same question about myself. I've added an
intentional failure in classlib, and then configured so that
mail_server=mail.optonline.net
from address, to address...
as I don't have smtp running on the test box, so I figured I could
just tell it where to go. It doesn't seem to be working. But I
don't see any bounces, nor do I see any moderator/admin messages from
the ASF mail infrastructure.
It would be nice if we had an ant target to to a test mail, becuase
waiting around for something to break is a *slow* process. I'm gonna
look at that if no one has a better idea.
geir
geir
On Jan 25, 2007, at 10:54 AM, Geir Magnusson Jr. wrote:
I think this test is bogus for two reasons.
1) As far as I can see, it's testing hashCode() of
junit.framework.TestCase
2) even if testing hashcode for a DateFormatSymbols object, I
have no idea why we'd assert a hascode has to be +ve
Comments? If no objections, I'm just going to modify the test
just to check for equality or even toss it
geir
