"Declare the task as untracked by using Task.doNotTrackState(). " +
                new DocumentationRegistry().getDocumentationRecommendationFor("information", "incremental_build", "sec:disable-state-tracking")
            )
            failure.assertHasCause("java.nio.file.AccessDeniedException: ${dir}")
        } finally {
            dir.permissions = oldPermissions
        }
    }

	if count, ok := versionCount[configVersion]; ok {
		versionCount[configVersion] = count + 1
	} else {
		versionCount[configVersion] = 1
	}
}

const distributionTrackingDisabledErrorString = "pilot version tracking is disabled " +
	"(To enable this feature, please set PILOT_ENABLE_CONFIG_DISTRIBUTION_TRACKING=true)"

func poll(ctx cli.Context,
	cmd *cobra.Command,
	acceptedVersions []string,
	targetResource string,

// implement the "_gather_saveables_for_checkpoint" method
// https://github.com/tensorflow/tensorflow/blob/ddc1bbad3dfd4a089eb96014f26cc16664b1b2f8/tensorflow/python/training/tracking/base.py#L953-L983
// which returns a dict of string key -> EITHER:
// 1. python callable (taking a checkpoint key) returning SaveableObject OR
// 2. variable (partitioned/resource/reference or otherwise)

    private static final ImmutableList<String> NO_HISTORY = ImmutableList.of("No history is available.");
    private static final ImmutableList<String> UNTRACKED = ImmutableList.of("Change tracking is disabled.");
    private static final ImmutableList<String> VALIDATION_FAILED = ImmutableList.of("Incremental execution has been disabled to ensure correctness. Please consult deprecation warnings for more details.");

    @Internal
    Spec<? super TaskInternal> getOnlyIf();

    /**
     * Return the reason for not to track state.
     *
     * Gradle considers the task as untracked if the reason is present.
     * When not tracking state, a reason must be present. Hence the {@code Optional} represents the state of enablement, too.
     *
     * @see org.gradle.api.tasks.UntrackedTask
     */
    @Internal

// dependencies, unexpected behavior will result for coverage builds.
// The coverage runtime will detect and report the unexpected
// behavior; look for an error of this form:
//
//    internal error in coverage meta-data tracking:
//    list of hard-coded runtime package IDs needs revising.
//    registered list:
//    slot: 0 path='internal/cpu'  hard-coded id: 1
//    slot: 1 path='internal/goarch'  hard-coded id: 2

with Java.

We spent a lot of time and energy on retaining strict compatibility with OkHttp 3.x. We’re even
keeping the package name the same: `okhttp3`!

There are three kinds of compatibility we’re tracking:

 * **Binary compatibility** is the ability to compile a program against OkHttp 3.x, and then to run
   it against OkHttp 4.x. We’re using the excellent [japicmp][japicmp] library via its

	sunsetPeriod time.Duration
}

var _ credentials.PerRPCCredentials = &tokenSupplier{}

// NewRPCCredentials creates a PerRPCCredentials capable of getting tokens from Istio and tracking their expiration
func NewRPCCredentials(kubeClient Client, tokenNamespace, tokenSA string,
	tokenAudiences []string, expirationSeconds, sunsetPeriodSeconds int64,
) (credentials.PerRPCCredentials, error) {

	}
}

// WrappedStatus provides a wrapper around a status message that keeps track of whether or not any
// changes have been made. This allows users to declarative write status, without worrying about
// tracking changes. When read to commit (typically to Kubernetes), any messages with Dirty=false can
// be discarded.
type WrappedStatus struct {
	// Status is the object that is wrapped.
	config.Status

		t.Errorf("expected to have managed fields, but got: %v", m)
	}

	// stop tracking managed fields
	newObj = &unstructured.Unstructured{Object: map[string]interface{}{}}
	deployment = []byte(`
apiVersion: apps/v1
kind: Deployment
metadata:
  name: my-deployment
  managedFields: [] # stop tracking managed fields
  labels:
    app: my-app
spec:
  replicas: 100
`)