When a **container** is started, it will run that command/program (although you can override it and make it run a different command/program).

A container is running as long as the **main process** (command or program) is running.

* It takes each **request** that comes to your application.
* It can then do something to that **request** or run any needed code.
* Then it passes the **request** to be processed by the rest of the application (by some *path operation*).

TensorFlow server (`tf.train.Server`). The TensorFlow server is intended for
internal communication only. It is not built for use in untrusted environments
or networks.

For performance reasons, the default TensorFlow server does not include any
authorization protocol and sends messages unencrypted. It accepts connections
from anywhere, and executes the graphs it is sent without performing any checks.

    // Write the mocking script.
    peer.sendFrame().settings(Settings())
    peer.acceptFrame() // ACK
    peer.sendFrame().ping(false, 2, 3)
    peer.acceptFrame() // PING
    peer.play()

    // Play it back.
    connect(peer)

    // Verify the peer received what was expected.
    val ping = peer.takeFrame()
    assertThat(ping.type).isEqualTo(Http2.TYPE_PING)
    assertThat(ping.streamId).isEqualTo(0)

}

// NewReader returns a new Reader that wraps src and computes
// MD5 checksum of everything it reads as ETag.
//
// It also computes the SHA256 checksum of everything it reads
// if sha256Hex is not the empty string.
//
// If size resp. actualSize is unknown at the time of calling
// NewReader then it should be set to -1.
// When size is >=0 it *must* match the amount of data provided by r.
//

   * rate at the end of the period (as long as there are enough requests to saturate it). Similarly,
   * if the {@code RateLimiter} is left <i>unused</i> for a duration of {@code warmupPeriod}, it
   * will gradually return to its "cold" state, i.e. it will go through the same warming up process
   * as when it was first created.
   *

But that's not the only way to declare dependencies (although it would probably be the more common).

The key factor is that a dependency should be a "callable".

A "**callable**" in Python is anything that Python can "call" like a function.

So, if you have an object `something` (that might _not_ be a function) and you can "call" it (execute it) like:

```Python
something()
```

or

```Python

        }
    }

    /**
     * Finds the parent symbol of the given [ResolveResult] by traversing back up the symbol hierarchy a [goBackSteps] steps,
     * or until the containing class or object symbol is found.
     *
     * Knows about the [ResolveResult.receiverClassReference] field and uses it in case it's not empty.
     */

   * contains both {@link Double#POSITIVE_INFINITY} and {@link Double#NEGATIVE_INFINITY} then the
   * result is {@link Double#NaN}. If it contains {@link Double#POSITIVE_INFINITY} and finite values
   * only or {@link Double#POSITIVE_INFINITY} only, the result is {@link Double#POSITIVE_INFINITY}.
   * If it contains {@link Double#NEGATIVE_INFINITY} and finite values only or {@link

  // This field cannot be empty for new Events and it can have at most 128 characters.
  optional string reportingInstance = 5;

  // action is what action was taken/failed regarding to the regarding object. It is machine-readable.
  // This field cannot be empty for new Events and it can have at most 128 characters.
  optional string action = 6;

  // reason is why the action was taken. It is human-readable.