final Supplier<T> delegate;
    transient volatile boolean initialized;
    // "value" does not need to be volatile; visibility piggy-backs
    // on volatile read of "initialized".
    @CheckForNull transient T value;

    MemoizingSupplier(Supplier<T> delegate) {
      this.delegate = checkNotNull(delegate);
    }

    @Override

   * cancelled without a call to {@code cancel}, such as by calling {@code
   * setFuture(cancelledFuture)}.
   *
   * <p>Beware of completing a future while holding a lock. Its listeners may do slow work or
   * acquire other locks, risking deadlocks.
   */
  @CanIgnoreReturnValue
  @Override
  public boolean cancel(boolean mayInterruptIfRunning) {
    Object localValue = value;
    boolean rValue = false;

    // is likely to be a bug anyway.

    // N.B. All writes to the list and the next pointers must have happened before the above
    // synchronized block, so we can iterate the list without the lock held here.
    RunnableExecutorPair reversedList = null;
    while (list != null) {
      RunnableExecutorPair tmp = list;
      list = list.next;
      tmp.next = reversedList;

   * Queue<E> queue = Queues.synchronizedQueue(MinMaxPriorityQueue.<E>create());
   * ...
   * queue.add(element);  // Needn't be in synchronized block
   * ...
   * synchronized (queue) {  // Must synchronize on queue!
   *   Iterator<E> i = queue.iterator(); // Must be in synchronized block
   *   while (i.hasNext()) {
   *     foo(i.next());
   *   }
   * }
   * }</pre>
   *

   * {@code tailSet} views.
   *
   * <pre>{@code
   * NavigableSet<E> set = synchronizedNavigableSet(new TreeSet<E>());
   *  ...
   * synchronized (set) {
   *   // Must be in the synchronized block
   *   Iterator<E> it = set.iterator();
   *   while (it.hasNext()) {
   *     foo(it.next());
   *   }
   * }
   * }</pre>
   *
   * <p>or:
   *
   * <pre>{@code

   * of reads may not be immediately reflected on the algorithm's data structures. These structures
   * are guarded by a lock and operations are applied in batches to avoid lock contention. The
   * penalty of applying the batches is spread across threads so that the amortized cost is slightly
   * higher than performing just the operation without enforcing the capacity constraint.

   * the {@code Future} is already done. Second, if buggy code calls {@code getDone} on a {@code
   * Future} that is still pending, the program will throw instead of block. This can be important
   * for APIs like {@link #whenAllComplete whenAllComplete(...)}{@code .}{@link
   * FutureCombiner#call(Callable, Executor) call(...)}, where it is easy to use a new input from

   *     HashMultimap.<K, V>create());
   * ...
   * Collection<V> values = multimap.get(key);  // Needn't be in synchronized block
   * ...
   * synchronized (multimap) {  // Synchronizing on multimap, not values!
   *   Iterator<V> i = values.iterator(); // Must be in synchronized block
   *   while (i.hasNext()) {
   *     foo(i.next());
   *   }
   * }
   * }</pre>
   *

 *   <li>If a {@code Throwable} is thrown in the try block, no exceptions that occur when attempting
 *       to close resources will be thrown from the finally block. The throwable from the try block
 *       will be thrown.
 *   <li>If no exceptions or errors were thrown in the try block, the <i>first</i> exception thrown
 *       by an attempt to close a resource will be thrown.

   * ...
   * Map<C, V> row = table.row(rowKey);  // Needn't be in synchronized block
   * ...
   * synchronized (table) {  // Synchronizing on table, not row!
   *   Iterator<Entry<C, V>> i = row.entrySet().iterator(); // Must be in synchronized block
   *   while (i.hasNext()) {
   *     foo(i.next());
   *   }
   * }
   * }</pre>
   *