VerifyFDefNodes(fdef, nodes);
    VerifyFDefInputs(fdef, inputs);
    VerifyFDefOutputs(fdef, outputs);
    VerifyFDefEdges(fdef, e_edges, c_edges, is_exact_edges);
  }

  // Serialize func_ to fdef and import it back
  void Reincarnate() {
    // func_ -> fdef
    tensorflow::FunctionDef fdef;
    ASSERT_TRUE(GetFunctionDef(func_, &fdef));
    TF_DeleteFunction(func_);

    // fdef -> func_

        * Scale subresource now has a new <code>targetSelector</code> field in its status. This field supports the new set-based selectors supported
by Deployments, but in a serialized format.
     * Spec change:
        * Deployment’s [selector](http://kubernetes.io/docs/user-guide/labels/#label-selectors) is now more general (supports set-based selector; it only supported
equality-based selector in 1.1).

 *
 * <p>The caches produced by {@code CacheBuilder} are serializable, and the deserialized caches
 * retain all the configuration properties of the original cache. Note that the serialized form does
 * <i>not</i> include cache contents, but only configuration.
 *
 * <p>See the Guava User Guide article on <a
 * href="https://github.com/google/guava/wiki/CachesExplained">caching</a> for a higher-level

				"application/vnd.android.package-archive",
				"application/x-tika-java-enterprise-archive",
				"application/x-tika-java-web-archive",
				"application/x-tika-unix-dump",
				"application/java-serialized-object",
				"application/javascript",
				"application/json",
				"application/java-vm",
				"application/x-java-jnilib",
				"application/kpml-request+xml",
				"application/kpml-response+xml",

*   Refactor kube-scheduler configuration ([#52428](https://github.com/kubernetes/kubernetes/pull/52428))
    *   The kube-scheduler command now supports a --config flag which is the location of a file containing a serialized scheduler configuration. Most other kube-scheduler flags are now deprecated. ([#52562](https://github.com/kubernetes/kubernetes/pull/52562),[ @ironcladlou](https://github.com/ironcladlou))

		//   inconsequential to the overall read operation.
		// - xl.meta metadata is still verified for quorum under lock()
		//   however writing the response doesn't need to serialize
		//   concurrent writers
		unlockOnDefer = true
		nsUnlocker = func() { lock.RUnlock(lkctx) }
	}

	fi, metaArr, onlineDisks, err := er.getObjectFileInfo(ctx, bucket, object, opts, true)
	if err != nil {

- The cloud config for Azure cloud provider can now be initialized from Kubernetes secret azure-cloud-provider in kube-system namespace
    - the secret is a serialized version of `azure.json` file with key cloud-config. And the secret name is azure-cloud-provider.

## Changes since v1.3.0-alpha.4

### Action Required

* Add direct serializer ([#26251](https://github.com/kubernetes/kubernetes/pull/26251), [@caesarxuchao](https://github.com/caesarxuchao))

    replaced with by writing regular Python functions. Using the `tf.function`
    decorator, such functions can be turned into graphs which can be executed
    remotely, serialized, and optimized for performance.
*   Unification of `tf.train.Optimizers` and `tf.keras.Optimizers`. Use
    `tf.keras.Optimizers` for TF2.0. `compute_gradients` is removed as public

  // Defaulted to nil.
  // +optional
  optional bool immutable = 5;

  // Data contains the secret data. Each key must consist of alphanumeric
  // characters, '-', '_' or '.'. The serialized form of the secret data is a
  // base64 encoded string, representing the arbitrary (possibly non-string)
  // data value here. Described in https://tools.ietf.org/html/rfc4648#section-4
  // +optional