// catch absolutely everything and fall through to our 'SafeAtomicHelper'
      // The access control checks that ARFU does means the caller class has to be AbstractFuture
      // instead of SafeAtomicHelper, so we annoyingly define these here
      try {
        helper =
            new SafeAtomicHelper(
                newUpdater(Waiter.class, Thread.class, "thread"),
                newUpdater(Waiter.class, Waiter.class, "next"),

       * template to be null (though we do handle that case gracefully at runtime). I've left this
       * one as it is because one of our users has defined a wrapper API around Preconditions,
       * declaring a checkState method that accepts a possibly null template. So we'd need to update
       * that user first.
       */
      @CheckForNull String errorMessageTemplate,

   * set of the empty set is not the empty set, but a one-element set containing the empty set.
   *
   * <p>The returned set and its constituent sets use {@code equals} to decide whether two elements
   * are identical, even if the input set uses a different concept of equivalence.
   *
   * <p><i>Performance notes:</i> while the power set of a set with size {@code n} is of size {@code

       * template to be null (though we do handle that case gracefully at runtime). I've left this
       * one as it is because one of our users has defined a wrapper API around Preconditions,
       * declaring a checkState method that accepts a possibly null template. So we'd need to update
       * that user first.
       */
      @CheckForNull String errorMessageTemplate,

 * loading), and fewer <a
 * href="https://github.com/google/guava/issues?q=is%3Aopen+is%3Aissue+label%3Apackage%3Dcache+label%3Atype%3Ddefect">bugs</a>.
 *
 * <p>Caffeine defines its own interfaces (<a
 * href="https://www.javadoc.io/doc/com.github.ben-manes.caffeine/caffeine/latest/com.github.benmanes.caffeine/com/github/benmanes/caffeine/cache/Cache.html">{@code
 * Cache}</a>, <a

         * element anymore. Otherwise, when we remove from the old iterator, we may be invalidating
         * the new one. The result is a ConcurrentModificationException or other bad behavior.
         *
         * (If we decide that we really, really hate allocating two Iterators per cycle instead of
         * one, we can optimistically store the new Iterator and then be willing to throw it out if
         * the user calls remove().)
         */