Credential tracking is intended to track relevant and valuable use of
the credential, not something as simple as using that credential in a
form validation.

Form validation is valuable for the user experience of the
administrator, but it is not valuable enough to track the credential
use in validating the form.

The comment is sufficient to explain why the implementation call
is not used.

…sing-null-context-to-credentials

JENKINS-44640 - Fix NPE checking legacy git SCM URL

Conflicts:
	pom.xml

Call User.getById when available

…ameterized trigger

Arguably we should bump to a newer version, but proposing just the minimal change to pick up jenkinsci/parameterized-trigger-plugin@440065d

It is 6 years since the version range dependency was removed from parameterized trigger

…nd toString implementation requirements

…lways `origin`

…lways `origin`

…scriptor

…aseline version of Jenkins core

… by GitSCMSource should honour refspecs on clone and not fetch tags

Conflicts:
	pom.xml

[JENKINS-26100] Switch buildEnvVars and others to Run

… message of exception

Use latest LTS (2.60.1) for plugin compat test

Passed a message when throwing AbortException so that it wouldn't show up as null

…s-output

[JENKINS-38241] Print commit message to log on checkout

…ethod names

…on for JENKINS-33445

…y fullName (the default behavior of the plugin).

Reverting jenkinsci#498 as it broke association of users to commits by fullName

[JENKINS-26100] Supply GitSCM/buildEnv

[JENKINS-43507] Allow SCMSource and SCMNavigator subtypes to share common traits

…he related code.

I was investigating the NPE stacktrace reported back in 2014, and it seems it is not possible anymore. Though I have added some API annotations to make the case explicitly verifyable in FindBugs.

The given_filesystem_when_askingChangesSinceNewRevision_then_changesArePopulatedButEmpty
test had a confusing mix of xxx261 and xxx260 variables that were
referring to tags which sometimes did not match the name of the variable.

Makes the names consistent rather than risk confusing the reader of the
test.

Other tests referred to a GitSCMFileSystem instance as "instance" when
it was then used as a file system reference to a specific commit.
Changed the name to be clearer where it was referring.

Preparing to remove the requirement that tests require all tags and
all history in the working directory.  Optimizations in git plugin
3.4.0 remove that assumption (and its resulting waste of time and
disc space).

…22593-annotations

[JENKINS-22593] - Annotate BuildChooser#getCandidateRevisions() and the related code

…Test-typos

Fix typos in GitSCMFileSystemTest naming

    <!-- promoted-builds requests 1.9.2, core requests 1.8.4 -->
    <exclude>org.apache.ant:ant</exclude>
    <!-- structs requests 1.9, core requests 1.7 -->
    <exclude>org.jenkins-ci:annotation-indexer</exclude>
    <!-- mockito requests 2.1, powermock requests 2.4 -->
    <exclude>org.objenesis:objenesis</exclude>

If -Djenkins.version=2.60.1 is used to compile, then a findbugs warning
is reported because the descriptor might be returned as null for the GitSCM
class.  The safeguard is harmless, and removes one findbugs warning from
the plugin compatibility test that is executed from the Jenkinsfile.

…parent-pom

Use parent pom 2.32, with a few exclusions

The method GitSCMSource.DescriptorImpl#doFillCredentialsIdItems
expects the parameter 'context' to be an SCMSourceOwner which breaks
the ability to add a global library at the level of a folder with the
modern SCM implementation. Changing the parameter type from
SCMSourceOwner to Item solves the issue.

…it plugin

With the update to git-client we can now fix the shallow clone issue on merge with

[FIXED JENKINS-45771] Consolidate the MergeWithGitSCMExtension into git plugin

[JENKINS-44751] Use ls-remote to determine live list of remote branches

…gin for PreBuildMerge Trait

[JENKINS-44271] doFillcredentialsId to look for Item not SCMSourceOwner.

…eOverrides

Added required overrides to PreBuildMerge extension

…ches.

… use