*
   * Past underutilization could mean that excess resources are available. Then, the RateLimiter
   * should speed up for a while, to take advantage of these resources. This is important when the
   * rate is applied to networking (limiting bandwidth), where past underutilization typically
   * translates to "almost empty buffers", which can be filled immediately.
   *

                // we dangerously assume E is a string
                List<E> extremeValues = (List<E>) getExtremeValues();
                @SuppressWarnings("unchecked")
                // map generators must past entry objects
                List<E> normalValues = (List<E>) Arrays.asList(entries);

                // prepare extreme values to be filtered out of view
                Collections.sort(extremeValues, comparator);

   *     arbitrary past or present time
   */
  abstract long queryEarliestAvailable(long nowMicros);

  /**
   * Reserves the requested number of permits and returns the time that those permits can be used
   * (with one caveat).
   *
   * @return the time that the permits may be used, or, if the permits may be used immediately, an
   *     arbitrary past or present time
   */

     *
     * The exception is the *InValueSet fields of multimapHeaderEntry, which are never set. (That
     * works out fine as long as we continue to be careful not to try to delete them or iterate
     * past them.)
     *
     * We could consider "lying" and omitting @CheckNotNull from all these fields. Normally, I'm not
     * a fan of that: What if we someday implement (presumably to be enabled during tests only)

     *
     * The exception is the *InValueSet fields of multimapHeaderEntry, which are never set. (That
     * works out fine as long as we continue to be careful not to try to delete them or iterate
     * past them.)
     *
     * We could consider "lying" and omitting @CheckNotNull from all these fields. Normally, I'm not
     * a fan of that: What if we someday implement (presumably to be enabled during tests only)

        if (lastFromForgetMeNot != null) {
          canRemove = true;
          return lastFromForgetMeNot;
        }
      }
      throw new NoSuchElementException("iterator moved past last element in queue.");
    }

    @Override
    public void remove() {
      checkRemove(canRemove);
      checkModCount();
      canRemove = false;
      expectedModCount++;

   *
   * <p>This is designed for generating persistent fingerprints of strings. It isn't
   * cryptographically secure, but it produces a high-quality hash with fewer collisions than some
   * alternatives we've used in the past.
   *
   * <p>FarmHash fingerprints are encoded by {@link HashCode#asBytes} in little-endian order. This
   * means {@link HashCode#asLong} is guaranteed to return the same value that

   *
   * Past underutilization could mean that excess resources are available. Then, the RateLimiter
   * should speed up for a while, to take advantage of these resources. This is important when the
   * rate is applied to networking (limiting bandwidth), where past underutilization typically
   * translates to "almost empty buffers", which can be filled immediately.
   *

       *
       * - in the "reentrant submit" case of `execute` (in which case we must have started running a
       *   user task -- which means that we already got past this code (or else we exited early
       *   above))
       */
      // Unconditionally set; there is no risk of throwing away a queued task from another thread,

 * addressable domain on the web -- that is, for a candidate string {@code "xxx"}, might browsing to
 * {@code "http://xxx/"} result in a webpage being displayed? In the past, this test was frequently
 * done by determining whether the domain ended with a {@linkplain #isPublicSuffix() public suffix}
 * but was not itself a public suffix. However, this test is no longer accurate. There are many