pages.

of project with maven-failsafe-plugin

Do not inject testFailureIgnore to
maven-failsafe-plugin:integration-test because it doesn't exist and
causes an NPE with Maven2.
Only insert it for maven-failsafe-plugin:verify.

"When running on Java 5, must still load DiskSpaceMonitor/TemporarySpaceMonitor, even though they will show N/A."

This reverts commit 8ad62f77.

[FIXED JENKINS-15869] When running on Java 5, must still load DiskSpaceMonitor/TemporarySpaceMonitor, even though they will show N/A.
Returning null from an @Extension method does not seem to work reliably, and is not supported by SezPoz.
If desirable to hide a monitor column in $server/computer/, add a boolean isEnabled() method interpretable in Jelly.

Wrap history iterable to hide class implementation details.
Jelly was getting confused by the history underlying class being implemented
as an extension of Map

[JENKINS-13202] 646154f1 refinement: resolveSymlink is documented to return null, not throw an exception, when called on a regular file.

The bug was actually already fixed and moreover this could cause that fireCompleted won't be called in case that cleanUp throws an exception, see commentes under https://github.com/jenkinsci/jenkins/commit/51f715162eba4dd1c4e081ec15c7ceb12e813ca5

Added link to the pull request of "Changed defaults for the Mac installer to make iOS codesigning easier"

[FIXED JENKINS-15733] ArrayIndexOutOfBoundsException from Launcher due to misinterpretation of ProcStarter.envs in f5cebdc8.
(Should this field be made a [LinkedHash]Map<String,String> or even EnvVars?
It is protected in a final class, so effectively package-private, thus not a b/w compat issue.)

maven-failsafe-plugin deserves the same level of integration
as maven-surefire-plugin.

A failed IT should mark the build as unstable, not failed.

Integrated the fix made in Winstone.

Integrated a fix made in Stapler.

This reverts commit c8e5cb93.