* wrapping a checked exception, prefer {@code ExecutionException}.
 *
 * @author Charles Fry
 * @since 10.0
 */
@GwtCompatible
public class UncheckedExecutionException extends RuntimeException {
  /*
   * Ideally, this class would have exposed only constructors that require a non-null cause. See

            throws ProjectBuildingException;

    // TODO this is only to provide a project for plugins that don't need a project to execute but need some
    // of the values from a MavenProject. Ideally this should be something internal and nothing outside Maven
    // would ever need this so it should not be exposed in a public API

         * This avoids out-of-control memory consumption, and it keeps the cache from growing so
         * large that doing the lookup is noticeably slower than redoing the work would be.
         *
         * Ideally we'd have a real eviction policy, but until we see a problem in practice, I hope
         * that this will suffice. I have not even benchmarked with different size limits.
         */
        if (validClasses.size() > 1000) {

By using all the previous alternatives I had the chance to learn from all of them, take ideas, and combine them in the best way I could find for myself and the teams of developers I have worked with.

For example, it was clear that ideally it should be based on standard Python type hints.

Also, the best approach was to use already existing standards.

 * continue to distinguish between exceptions and errors, even when they come from other threads.
 *
 * @author Chris Povirk
 * @since 10.0
 */
@GwtCompatible
public class ExecutionError extends Error {
  /*
   * Ideally, this class would have exposed only constructors that require a non-null cause. See

        // respectively. Using a simple replacement of these patterns will allow them
        // to specify the absolute path to these files in a customized components.xml
        // file. Ideally, we'd do full pattern-evaluation against the sysprops, but this
        // is a first step. There are several replacements below, in order to normalize
        // the path character before we operate on the string as a regex input, and

    putEdge(N2, N1);
    assertThat(graph.outDegree(N1)).isEqualTo(3);
  }

  // Stable order tests

  // Note: Stable order means that the ordering doesn't change between iterations and versions.
  // Ideally, the ordering in test should never be updated.
  @Test
  public void stableIncidentEdgeOrder_edges_returnsInStableOrder() {
    assume().that(graph.incidentEdgeOrder().type()).isEqualTo(ElementOrder.Type.STABLE);

    putEdge(N1, N2);
    assertThat(graph.outDegree(N1)).isEqualTo(2);
  }

  // Stable order tests

  // Note: Stable order means that the ordering doesn't change between iterations and versions.
  // Ideally, the ordering in test should never be updated.
  @Test
  public void stableIncidentEdgeOrder_edges_returnsInStableOrder() {
    assume().that(graph.incidentEdgeOrder().type()).isEqualTo(ElementOrder.Type.STABLE);

         * This avoids out-of-control memory consumption, and it keeps the cache from growing so
         * large that doing the lookup is noticeably slower than redoing the work would be.
         *
         * Ideally we'd have a real eviction policy, but until we see a problem in practice, I hope
         * that this will suffice. I have not even benchmarked with different size limits.
         */
        if (validClasses.size() > 1000) {

   * without contention. Because assignment of entries to these partitions is not necessarily
   * uniform, the actual concurrency observed may vary. Ideally, you should choose a value to
   * accommodate as many threads as will ever concurrently modify the table. Using a significantly
   * higher value than you need can waste space and time, and a significantly lower value can lead