// stream is closed (we won't use any further bytes) and the output stream is either finished
      // or closed (so RSTing both streams doesn't cause harm).
      ******@****.***(ErrorCode.CANCEL, null)
    } else if (!open) {
      connection.removeStream(id)
    }
  }

   * world. Figure out what sort of space-time tradeoff we're actually going to get here with the
   * *Map variants. This class is particularly hard to benchmark, because the benefit is not only in
   * less allocation, but also having the GC do less work to scan the heap because of fewer
   * references, which is particularly hard to quantify.
   */

  /** Creates an empty {@code CompactHashMap} instance. */

    handler: Handler,
    length: Int,
    flags: Int,
    streamId: Int,
  ) {
    if (streamId == 0) throw IOException("PROTOCOL_ERROR: TYPE_DATA streamId == 0")

    // TODO: checkState open or half-closed (local) or raise STREAM_CLOSED
    val inFinished = flags and FLAG_END_STREAM != 0
    val gzipped = flags and FLAG_COMPRESSED != 0
    if (gzipped) {

  }

  /*
   * Whenever an implementation uses `instanceof` on a parameter instance, the test has to know that
   * (so much for "black box") and try instances that both do and don't pass the check. The "don't"
   * half of that is more awkward to arrange...
   */
  private static <T> Iterable<T> iterable(Collection<T> collection) {
    // return collection::iterator;
    return new Iterable<T>() {
      @Override

  }

  /*
   * Whenever an implementation uses `instanceof` on a parameter instance, the test has to know that
   * (so much for "black box") and try instances that both do and don't pass the check. The "don't"
   * half of that is more awkward to arrange...
   */
  private static <T> Iterable<T> iterable(Collection<T> collection) {
    // return collection::iterator;
    return new Iterable<T>() {
      @Override

   *
   * <p><b>Performance note:</b> The resulting stream is not <a
   * href="http://gee.cs.oswego.edu/dl/html/StreamParallelGuidance.html">efficiently splittable</a>.
   * This may harm parallel performance.
   */
  @Beta
  public static <A extends @Nullable Object, B extends @Nullable Object, R extends @Nullable Object>
      Stream<R> zip(

e{content:"\f255"}.fa-hand-scissors:before{content:"\f257"}.fa-hand-spock:before{content:"\f259"}.fa-hands:before{content:"\f4c2"}.fa-hands-helping:before{content:"\f4c4"}.fa-handshake:before{content:"\f2b5"}.fa-hanukiah:before{content:"\f6e6"}.fa-hard-hat:before{content:"\f807"}.fa-hashtag:before{content:"\f292"}.fa-hat-cowboy:before{content:"\f8c0"}.fa-hat-cowboy-side:before{content:"\f8c1"}.fa-hat-wizard:before{content:"\f6e8"}.fa-hdd:before{content:"\f0a0"}.fa-heading:before{content:"\f1dc"}...

    val data = ByteArray(1024 * 1024)
    var i = 0
    while (i < 1024) {
      try {
        out!!.write(data)
        out.flush()
        if (i == 513) {
          // pause slightly after half way to make result more predictable
          Thread.sleep(100)
        }
      } catch (e: IOException) {
        break
      }
      i++
    }
    // Halfway +/- 0.5%

   * most key spaces are ANYTHING BUT uniformly distributed. A bit(i) in the input is said to
   * 'affect' a bit(j) in the output if two inputs, identical but for bit(i), will differ at output
   * bit(j) about half the time
   *
   * <p>Funneling is pretty simple to detect. The key idea is to find example keys which
   * unequivocally demonstrate that funneling cannot be occurring. This is done bit-by-bit. For each

          // have no idea.
          // TODO(lukes): consider building everything in terms of ListenableScheduledFuture then
          // the AbstractService could monitor the future directly. Rescheduling is still hard...
          // but it would help with some of these lock ordering issues.
          scheduleFailure = e;
          toReturn = new FutureAsCancellable(immediateCancelledFuture());
        } finally {