* `getDestinationDir()` was replaced by `getDestinationDirectory()`.
 * `getBinaryFile()`, `getOutputFile()` was replaced by `getLinkedFile()`.
 * `setOutputFile(File)` was removed. Use `Property.set()` instead.
 * `setOutputFile(Provider)` was removed. Use `Property.set()` instead.
 * `getTargetPlatform()` was changed to return a `Property`.
 * `setTargetPlatform(NativePlatform)` was removed. Use `Property.set()` instead.

    exception, "Expected Android API level 21+ but was 29".


## Version 3.14.6

_2020-01-11_

 *  Fix: Don't crash if the connection is closed when sending a degraded ping. This fixes a
    regression that was introduced in OkHttp 3.14.5.


## Version 3.14.5

_2020-01-03_

   * @throws IllegalStateException if a maximum size or weight was already set
   */
  @CanIgnoreReturnValue
  public CacheBuilder<K, V> maximumSize(long maximumSize) {
    checkState(
        this.maximumSize == UNSET_INT, "maximum size was already set to %s", this.maximumSize);
    checkState(
        this.maximumWeight == UNSET_INT,
        "maximum weight was already set to %s",
        this.maximumWeight);

The retention period was configured using `buildCache.local.removeUnusedEntriesAfterDays`.

In Gradle 8.0, link:directory_layout.html#dir:gradle_user_home:configure_cache_cleanup[a new mechanism] was added to configure the cleanup and retention periods for various resources in Gradle User Home.

=== IncrementalTaskInputs type is deprecated

The `IncrementalTaskInputs` type was used to implement _incremental tasks,_ that is to say tasks that can be optimized to run on a subset of changed inputs instead of the whole input.
This type had a number of drawbacks.
In particular using this type it was not possible to determine what input a change was associated with.

You should now use the `InputChanges` type instead.

	// and we haven't observed that it was stolen from us yet.
	state, ok := o.pStates[pid]
	if !ok || state.status != go122.ProcIdle || !seq.succeeds(state.seq) || curCtx.P != NoProc {
		// We can't make an inference as to whether this is bad. We could just be seeing
		// a ProcStart on a different M before the proc's state was emitted, or before we
		// got to the right point in the trace.
		//

        failure.assertHasCause("Could not resolve project :b.")
        failure.assertHasCause("""No matching variant of project :b was found. The consumer was configured to find attribute 'buildType' with value 'debug', attribute 'flavor' with value 'free' but:
  - Variant 'bar':

// shuttle.
//
// Originally, the work of this controller was split amongst three
// controllers. This controller is the result a large effort to simplify the
// PV subsystem. During that effort, it became clear that we needed to ensure
// that every single condition was handled and accounted for in the code, even
// if it resulted in no-op code branches.
//

		logger.V(6).Info("claim with structured parameters got deallocated", "pod", klog.KObj(pod), "claim", klog.KObj(modifiedClaim))
		return framework.Queue, nil
	}

	if !usesClaim {
		// This was not the claim the pod was waiting for.
		logger.V(6).Info("unrelated claim got modified", "pod", klog.KObj(pod), "claim", klog.KObj(modifiedClaim))
		return framework.QueueSkip, nil
	}

	if originalClaim == nil {

            'org:test:1.0' {
                expectGetMetadata()
            }
        }
        fails 'checkDeps'

        then:
        failure.assertHasCause("""No matching variant of org:test:1.0 was found. The consumer was configured to find a component for use during runtime, as well as attribute 'custom' with value 'c1' but:
  - Variant 'api' declares a component, as well as attribute 'custom' with value 'c1':