GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE
USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF
DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD
PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS),
EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF
SUCH DAMAGES.

- Fixed authentication_duration_seconds metric scope. Previously, it included whole apiserver request duration which yields inaccurate results. ([#99944](https://github.com/kubernetes/kubernetes/pull/99944), [@marseel](https://github.com/marseel))

* Disable per-cpu metrics by default for scalability. ([#60106](https://github.com/kubernetes/kubernetes/pull/60106), [@dashpole](https://github.com/dashpole))
    * Fix inaccurate disk usage monitoring of overlayFs.
    * Retry docker connection on startup timeout to avoid permanent loss of metrics.

import java.lang.annotation.RetentionPolicy;
import java.lang.annotation.Target;

/**
 * Outer class that exists solely to let us write {@code Partially.GwtIncompatible} instead of plain
 * {@code GwtIncompatible}. This is more accurate for {@link Futures#catching}, which is available
 * under GWT but with a slightly different signature.
 *
 * <p>We can't use {@code PartiallyGwtIncompatible} because then the GWT compiler wouldn't recognize

      <outputDirectory>/</outputDirectory>
      <includes>
        <include>DEPENDENCIES</include>
        <!-- exclude the license and notice as they are not as accurate as the ones from above -->
      </includes>
    </fileSet>
  </fileSets>

  }

  @CollectionFeature.Require(absent = SUPPORTS_ADD)
  @CollectionSize.Require(absent = ZERO)
  /*
   * absent = ZERO isn't required, since unmodList.add() must
   * throw regardless, but it keeps the method name accurate.
   */
  public void testAdd_unsupportedPresent() {
    assertThrows(UnsupportedOperationException.class, () -> getList().add(e0()));
  }

  @CollectionFeature.Require(value = {SUPPORTS_ADD, ALLOWS_NULL_VALUES})

  }

  @CollectionFeature.Require(absent = SUPPORTS_ADD)
  @CollectionSize.Require(absent = ZERO)
  /*
   * absent = ZERO isn't required, since unmodList.add() must
   * throw regardless, but it keeps the method name accurate.
   */
  public void testAdd_unsupportedPresent() {
    assertThrows(UnsupportedOperationException.class, () -> getList().add(e0()));
  }

  @CollectionFeature.Require(value = {SUPPORTS_ADD, ALLOWS_NULL_VALUES})

  public void decrement() {
    add(-1L);
  }

  /**
   * Returns the current sum. The returned value is NOT an atomic snapshot; invocation in
   * the absence of concurrent updates returns an accurate result, but concurrent updates that occur
   * while the sum is being calculated might not be incorporated.
   *
   * @return the sum
   */
  @Override
  public long sum() {
    long sum = base;

  }

  @ListFeature.Require(absent = SUPPORTS_ADD_WITH_INDEX)
  @CollectionSize.Require(absent = ZERO)
  /*
   * absent = ZERO isn't required, since unmodList.add() must
   * throw regardless, but it keeps the method name accurate.
   */
  public void testAddAtIndex_unsupportedPresent() {
    assertThrows(UnsupportedOperationException.class, () -> getList().add(0, e0()));
    expectUnchanged();
  }

Do not make a parameter `@Nullable` just to check and throw a `NullPointerException` when it is `null`. 

## Consequences

* Classes are forced to have consistent nullability annotations.
* IDE warnings become more accurate, reducing warnings fatigue.
* Public API nullability annotations become closer to reality.
* Java compilation of Gradle code suffers from small performance penalty (up to 10%).