Specifically, to see Uvicorn, Starlette and FastAPI compared together (among many other tools).

The simpler the problem solved by the tool, the better performance it will get. And most of the benchmarks don't test the additional features provided by the tool.

The hierarchy is like:

* **Uvicorn**: an ASGI server
    * **Starlette**: (uses Uvicorn) a web microframework

    // unless it's actually going to be used; validate() is called for all set method calls, so it
    // needs to be fast.
    // (We could instead generate the message once, when the set is created, but zero is better.)
    if (!validator.get()) {
      throw new IllegalStateException(errorMessage.get());
    }
  }

    id: what-did-you-do
    attributes:
      label: "What did you do?"
      description: "If possible, provide a recipe for reproducing the error. A complete runnable program is good. A link on [go.dev/play](https://go.dev/play) is better. A failing unit test is the best."
    validations:
      required: true
  - type: textarea
    id: actual-behavior
    attributes:
      label: "What did you see happen?"
    validations:
      required: true

# Reference

Here's the reference or code API, the classes, functions, parameters, attributes, and
all the FastAPI parts you can use in your applications.

If you want to **learn FastAPI** you are much better off reading the

import org.jspecify.annotations.NullUnmarked;

/**
 * Utility class for being able to seed a {@link Random} value with a passed in seed from a
 * benchmark parameter.
 *
 * <p>TODO: Remove this class once Caliper has a better way.
 *
 * @author Nicholaus Shupe
 */
@NullUnmarked
public final class SpecialRandom extends Random {
  public static SpecialRandom valueOf(String s) {

<a name="http2_naughty"></a>
#### ² HTTP/2 Cipher Suite Denylist

Cipher suites that are [discouraged for use][http2_denylist] with HTTP/2. OkHttp includes them because better suites are not commonly available. For example, none of the better cipher suites listed above shipped with Android 4.4 or Java 7.

[OkHttp30]: https://square.github.io/okhttp/changelog_3x/#version-300

/**
 * Utility class for handling Base64 encoding and decoding.
 * <p>
 * This class now uses the standard {@link java.util.Base64} implementation
 * instead of a custom implementation, providing better security and performance.
 * The API remains backward compatible with previous versions.
 * </p>
 *
 * @author higa
 */
public abstract class Base64Util {

    /**
     * Do not instantiate.
     */

### Is concurrency better than parallelism? { #is-concurrency-better-than-parallelism }

Nope! That's not the moral of the story.

Concurrency is different than parallelism. And it is better on **specific** scenarios that involve a lot of waiting. Because of that, it generally is a lot better than parallelism for web application development. But not for everything.

# The Gradle tests need the pom.xml only to read its version number.
# (And the file needs to be two directory levels up from the Gradle build file.)
# TODO(cpovirk): Find a better way to give them that information.
cp pom.xml "${GRADLE_TEMP}"

for version in 5.6.4 7.0.2; do
  # Enter a subshell so that we return to the current directory afterward.
  (
    cp -r integration-tests "${GRADLE_TEMP}/${version}"
    cd "${GRADLE_TEMP}/${version}/gradle"

 * under the License.
 */
package org.apache.maven.artifact.resolver.filter;

import org.apache.maven.artifact.Artifact;

/**
 * Filter to only retain objects in the given artifactScope or better.
 *
 */
abstract class AbstractScopeArtifactFilter implements ArtifactFilter {

    private boolean compileScope;

    private boolean runtimeScope;

    private boolean testScope;