* For example, resolves "[3.8,4.0)" to "3.8", "3.8.1", "3.8.2".
     * The returned list of versions is only dependent on the configured repositories and their contents.
     * The supplied request may also refer to a single concrete version rather than a version range.
     * In this case though, the result contains simply the (parsed) input version, regardless of the
     * repositories and their contents.
     *

and replaced with those of `initial`. From the repo root:

    > cd doc
    > rm -r next/*
    > cp -r initial/* next

Then edit `next/1-intro.md` to refer to the next version.

To prepare the release notes for a release, run `golang.org/x/build/cmd/relnote generate`.

 * is meant for exclusive consumption by the model builder and is opaque to the cache implementation. The cache key is
 * formed by a combination of group id, artifact id, version and tag. The first three components generally refer to the
 * identity of a model. The tag allows for further classification of the associated data on the sole discretion of the
 * model builder.
 *
 */
public interface ModelCache {

//revive:disable:var-naming Underscores is used here to indicate where common prefix ends and the enumeration name begins
const (
	lcEventSrc_None lcEventSrc = iota
	lcEventSrc_Heal
	lcEventSrc_Scanner
	lcEventSrc_Decom
	lcEventSrc_Rebal
	lcEventSrc_s3HeadObject
	lcEventSrc_s3GetObject
	lcEventSrc_s3ListObjects
	lcEventSrc_s3PutObject
	lcEventSrc_s3CopyObject
	lcEventSrc_s3CompleteMultipartUpload
)

//revive:enable:var-naming

			expectedString: `"deployment/random-gibberish" does not refer to a pod`,
			wantException:  true,
		},
		{ // supplying nonexistent deployment name in nonexistent namespace
			args:           strings.Split("endpoint deployment/random-gibberish.bogus", " "),
			expectedString: `"deployment/random-gibberish" does not refer to a pod`,
			wantException:  true,
		},

 * is meant for exclusive consumption by the model builder and is opaque to the cache implementation. The cache key is
 * formed by a combination of group id, artifact id, version and tag. The first three components generally refer to the
 * identity of a model. The tag allows for further classification of the associated data on the sole discretion of the
 * model builder.
 *
 */
public interface ModelCache {

    ): List<SmallSubprojectBucket> {
        // splitIntoBuckets() method expects us to split large element into N elements,
        // but we want to have a single bucket with N batches.
        // As a workaround, we repeat the bucket N times, and deduplicate the result at the end
        val resultIncludingDuplicates = splitIntoBuckets(
            LinkedList(subProjectTestClassTimes),
            SubprojectTestClassTime::totalTime,

Packages are imported one by one in alphabetical order during runtime.

The test doesn't identify package's order-dependent issues; for instance,
importing "tf.foo" followed by "tf.bar" won't reveal that "tf.bar" depends on
"tf.foo" being imported first.

The `_api/v2/api_packages.txt` file is generated during the TensorFlow API v2
init files creation process and is subsequently stored in the wheel file after

	// Re-run the stringer command to generate them again.
	var x [1]struct{}
	_ = x[lcEventSrc_None-0]
	_ = x[lcEventSrc_Heal-1]
	_ = x[lcEventSrc_Scanner-2]
	_ = x[lcEventSrc_Decom-3]
	_ = x[lcEventSrc_Rebal-4]
	_ = x[lcEventSrc_s3HeadObject-5]
	_ = x[lcEventSrc_s3GetObject-6]
	_ = x[lcEventSrc_s3ListObjects-7]
	_ = x[lcEventSrc_s3PutObject-8]
	_ = x[lcEventSrc_s3CopyObject-9]
	_ = x[lcEventSrc_s3CompleteMultipartUpload-10]

	// doesn't really matter anyway, since this is purely a string
	// function we're testing, and it's not actually being used to
	// do a system call)
	veryLong := "l" + strings.Repeat("o", 248) + "ng"
	for _, test := range []struct{ in, want string }{
		// Short; unchanged:
		{`C:\short.txt`, `C:\short.txt`},
		{`C:\`, `C:\`},
		{`C:`, `C:`},
		// The "long" substring is replaced by a looooooong