// and return value flows unmodified into an indirect call)
//
// For categories 2 and 3 above, each adjustment can have either a
// "must" version and a "may" version (but not both). Here the idea is
// that in the "must" version the value flow is unconditional: if the
// callsite executes, then the condition we're interested in (ex:
// param feeding call) is guaranteed to happen. For the "may" version,

func (hs *clientHandshakeStateTLS13) processHelloRetryRequest() error {
	c := hs.c

	// The first ClientHello gets double-hashed into the transcript upon a
	// HelloRetryRequest. (The idea is that the server might offload transcript
	// storage to the client in the cookie.) See RFC 8446, Section 4.4.1.
	chHash := hs.transcript.Sum(nil)
	hs.transcript.Reset()

    * Automatic client code generation systems, for many languages.
* Provide 2 interactive documentation web interfaces directly.

---

We just scratched the surface, but you already get the idea of how it all works.

Try changing the line with:

```Python
    return {"item_name": item.name, "item_id": item_id}
```

...from:

```Python
        ... "item_name": item.name ...
```

// updating the in-memory copy of the latest tree head (c.latest)
// not the configuration file.
// The when result explains when msg happened relative to our
// previous idea of c.latest:
// msgPast means msg was from before c.latest,
// msgNow means msg was exactly c.latest, and
// msgFuture means msg was from after c.latest, which has now been updated.

	if err != nil {
		// We cannot parse it - must be some new schema we don't know about. We should assume we do not manage it.
		// In theory, this should never happen, unless we decide a number was a bad idea in the future.
		return cur, false, false
	}
	if curNum > ControllerVersion {
		// A newer version owns this gateway, let them handle it
		return cur, false, false
	}
	if curNum == ControllerVersion {

 *       check for thread assertion failures.
 *   <li>All delays and timeouts must use one of the constants {@code SHORT_DELAY_MS}, {@code
 *       SMALL_DELAY_MS}, {@code MEDIUM_DELAY_MS}, {@code LONG_DELAY_MS}. The idea here is that a
 *       SHORT is always discriminable from zero time, and always allows enough time for the small
 *       amounts of computation (creating a thread, calling a few methods, etc) needed to reach a

You can define an automatic module as part of a multi-project that otherwise defines real modules (e.g. as an adapter to another library).
While this works fine in the Gradle build, such automatic module projects are not correctly recognized by IDEA/Eclipse at the moment.
You can work around it by manually adding the Jar built for the automatic module to the dependencies of the project that does not find it in the IDE's UI.
===

[[sec:java_library_classes_usage]]

// search process down to just two DNS queries. This will eliminate unnecessary upstream DNS queries from the
// agent, reduce load on DNS servers and improve overall latency. This idea was borrowed and adapted from
// the autopath plugin in coredns. The implementation here is very different from auto path though.
// Autopath does inline computation to see if the given query could potentially match something else

### Concurrency and Burgers

This idea of **asynchronous** code described above is also sometimes called **"concurrency"**. It is different from **"parallelism"**.

**Concurrency** and **parallelism** both relate to "different things happening more or less at the same time".

 *       check for thread assertion failures.
 *   <li>All delays and timeouts must use one of the constants {@code SHORT_DELAY_MS}, {@code
 *       SMALL_DELAY_MS}, {@code MEDIUM_DELAY_MS}, {@code LONG_DELAY_MS}. The idea here is that a
 *       SHORT is always discriminable from zero time, and always allows enough time for the small
 *       amounts of computation (creating a thread, calling a few methods, etc) needed to reach a