Regular Reads will block until data is available, but not wait for a full buffer. 
Writes will block until there is space available and writes bigger than the buffer will wait for reads to make space.

`TryRead` and `TryWrite` are still available for non-blocking reads and writes.

/**
 * Reads up to b.length bytes of data from this input stream into an array of bytes.
 *
 * @throws IOException if a network error occurs
 */

    public int read( byte[] b ) throws IOException {
        return read( b, 0, b.length );
    }

/**
 * Reads up to len bytes of data from this input stream into an array of bytes.
 *
 * @throws IOException if a network error occurs
 */

     * @throws IOException If the metadata could not be deserialized.
     * @throws MetadataParseException If the input format could not be parsed.
     */
    Metadata read(File input, Map<String, ?> options) throws IOException, MetadataParseException;

    /**
     * Reads the metadata from the specified character reader. The reader will be automatically closed before the method
     * returns.
     *

     * @throws IOException If the settings could not be deserialized.
     * @throws SettingsParseException If the input format could not be parsed.
     */
    Settings read(File input, Map<String, ?> options) throws IOException, SettingsParseException;

    /**
     * Reads the settings from the specified character reader. The reader will be automatically closed before the method
     * returns.
     *

     * @throws ToolchainsParseException If the input format could not be parsed.
     */
    PersistedToolchains read(File input, Map<String, ?> options) throws IOException, ToolchainsParseException;

    /**
     * Reads the toolchains from the specified character reader. The reader will be automatically closed before the
     * method returns.
     *

Framing rules make it impractical to implement http/2 correctly on a single blocking thread. The flow-control features introduce feedback between reads and writes, requiring writes to acknowledge reads and reads to throttle writes.

In OkHttp we expose a blocking API over a framed protocol. This document explains the code and policy that makes that work.

### Threads

#### Application's calling thread

      current = it.next().openStream();
    }
  }

  @Override
  public int read(char[] cbuf, int off, int len) throws IOException {
    checkNotNull(cbuf);
    if (current == null) {
      return -1;
    }
    int result = current.read(cbuf, off, len);
    if (result == -1) {
      advance();
      return read(cbuf, off, len);
    }
    return result;
  }

  @Override

    @Throws(IOException::class)
    override fun read(
      sink: Buffer,
      byteCount: Long,
    ): Long {
      while (left == 0) {
        source.skip(padding.toLong())
        padding = 0
        if (flags and FLAG_END_HEADERS != 0) return -1L
        readContinuationHeader()
        // TODO: test case for empty continuation header?
      }

      val read = source.read(sink, minOf(byteCount, left.toLong()))

    return (int) in.skip(n);
  }

  @CanIgnoreReturnValue // to skip a byte
  @Override
  public int readUnsignedByte() throws IOException {
    int b1 = in.read();
    if (0 > b1) {
      throw new EOFException();
    }

    return b1;
  }

  /**
   * Reads an unsigned {@code short} as specified by {@link DataInputStream#readUnsignedShort()},
   * except using little-endian byte order.
   *

TF_CAPI_EXPORT extern TFE_MonitoringCounterReader*
TFE_MonitoringNewCounterReader(const char* name);

// Reads the value of a counter that was created with 0 labels.
TF_CAPI_EXPORT extern int64_t TFE_MonitoringReadCounter0(
    TFE_MonitoringCounterReader*);

// Reads the value of specific cell of a counter that was created with 1 label.
TF_CAPI_EXPORT extern int64_t TFE_MonitoringReadCounter1(