On Jul 17, 2006, at 5:12 PM, David Jencks wrote:
After I wrote the tests I made many changes to make them pass on
more machines than just mine. I'd love to know how to make them
independent of machine load/speed. Do you have any concrete
suggestions on how to do this?
At the moment
I did not turn them off... I just added a Jira to remind us that
these tests need to be fixed.
--jason
On Jul 14, 2006, at 8:43 PM, Prasad Kashyap wrote:
The timer tests failed. Jason, I thought you pulled them out late last
night after our discussion.
Cheers
Prasad
On 7/14/06, Prasad
IIRC these tests have been running successfully (in M1) for more
than a year on all kinds of machines.
Thanks
Anita
--- Jason Dillon [EMAIL PROTECTED] wrote:
I did not turn them off... I just added a Jira to remind us that
these tests need to be fixed.
--jason
On Jul 14, 2006, at
Hi Prasad,
I wish I could be of more help and investigate why the config are
not building. These patches used to work until rev 418907 on the trunk.
I have been unable to build the trunk using 'mvn' after RTC-2161. I
have not had time to investigate it. IIRC this error comes when the
m1, m2 or ant... the timer tests are not written in such a way that
they will always pass. They are heavily dependent on running on a
healthy/speedy box or they will start to fail with random errors.
These tests run fine for me, but fail for Prasad. The failures are
all due to
Prasad, could you peek at the patches and help sort out what needs to
get applied to m2migration?
Thanks,
--jason
On Jul 15, 2006, at 4:24 AM, anita kulshreshtha wrote:
Hi Prasad,
I wish I could be of more help and investigate why the config are
not building. These patches used to
It looks like the Maven 2 peeps are not going to be able to take in
our patches to apply to the assembly plugin, which provides mappers
and some includes/excludes to unpack and a permission related fix to
files/file element processing.
I believe that those changes will eventually make it
Would it not be simpler to use the patched version of m-a-p and just
rename it to geronimo-assembly2-plugin and add it to our code base?
This plugin is more elegant than the ant solution. we can remove this
code as soon as maven-assembly-plugin is updated. Are there any
problems with using code
On Jul 14, 2006, at 4:30 PM, anita kulshreshtha wrote:
Would it not be simpler to use the patched version of m-a-p and
just
rename it to geronimo-assembly2-plugin and add it to our code base?
I do not believe so. The use of the dependency + antrun plugins are
relatively limited.
Both
I tried the latest changes on Windows and it builds the assembly fine.
I even ran the jetty binary successfully.
Jason, I also tried commenting out a few builders from the
dependencies list and it still worked fine. So I guess we can prune
the deps list to remove the redudant ones.
A full
The bootstrap on linux first failed a test in the security module. I
ran the build inside the security module. It passed successfully.
Then I ran bootstrap again. It failed while transferring an artifact.
Now I'm running mvn from top level. This is running.
Keeping my fingers crossed
Prasad
The timer tests failed. Jason, I thought you pulled them out late last
night after our discussion.
Cheers
Prasad
On 7/14/06, Prasad Kashyap [EMAIL PROTECTED] wrote:
The bootstrap on linux first failed a test in the security module. I
ran the build inside the security module. It passed
12 matches
Mail list logo
