Summary:      healNotStartedStatus,
			HealSettings: hs,
		},
		traverseAndHealDoneCh: make(chan error),
		cancelCtx:             cancel,
		ctx:                   ctx,
		scannedItemsMap:       make(map[madmin.HealItemType]int64),
		healedItemsMap:        make(map[madmin.HealItemType]int64),
		healFailedItemsMap:    make(map[madmin.HealItemType]int64),
	}
}

		}
		if i > 0 {
			if wait := globalBatchConfig.ExpirationWait(); wait > 0 {
				time.Sleep(wait)
			}
		}
		i++
		wk.Take()
		go func(toExpire []expireObjInfo) {
			defer wk.Give()

			toExpireAll := make([]expireObjInfo, 0, len(toExpire))
			toDel := make([]ObjectToDelete, 0, len(toExpire))
			oiCache := newObjInfoCache()
			for _, exp := range toExpire {
				if exp.ExpireAll {

Sometimes because some classes have lots of dependencies that make them hard to
instantiate. Sometimes because API contracts make tests hard to write. Code
refactors that make functionality easier to unit test are encouraged. If this
sounds very abstract to you, you can have a look at
https://github.com/elastic/elasticsearch/pull/16610[this pull request] for

	totalShards := len(partsMetadata)
	parities = make([]int, len(partsMetadata))
	for index, metadata := range partsMetadata {
		if errs[index] != nil {
			parities[index] = -1
			continue
		}
		if !metadata.IsValid() {
			parities[index] = -1
			continue
		}
		//nolint:gocritic
		// Delete marker or zero byte objects take highest parity.
		if metadata.Deleted || metadata.Size == 0 {

            classes that are actually used in the module under test.

            One way to eliminate the warnings is to make base.testModule include the not really
            necessary <inherits> lines for c.g.c.collect.testModule, etc. However, adding <inherits>
            lines could make c.g.c.base.testModule transitively inherit from extra modules. If some

	if p.dontSave {
		return nil
	}
	data := make([]byte, 4, p.Msgsize()+4)

	// Initialize the header.
	binary.LittleEndian.PutUint16(data[0:2], poolMetaFormat)
	binary.LittleEndian.PutUint16(data[2:4], poolMetaVersion)

	buf, err := p.MarshalMsg(data)
	if err != nil {
		return err
	}

	// Saves on all pools to make sure decommissioning of first pool is allowed.
	for i, eset := range pools {

	localDrives := cloneDrives(globalLocalDrivesMap)
	globalLocalDrivesMu.RUnlock()

	errs := make([]error, len(localDrives))
	infos := make([]DiskInfo, len(localDrives))
	for i, drive := range localDrives {
		infos[i], errs[i] = drive.DiskInfo(GlobalContext, DiskInfoOptions{})
	}
	infoMaps := make(map[string]madmin.DiskMetrics)
	for i := range infos {

## Restarts { #restarts }

Similar to making sure your application is run on startup, you probably also want to make sure it is **restarted** after failures.

### We Make Mistakes { #we-make-mistakes }

We, as humans, make **mistakes**, all the time. Software almost *always* has **bugs** hidden in different places. 🐛

</blockquote>

## Investigation { #investigation }

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.

///

## OpenID Connect { #openid-connect }

OpenID Connect is another specification, based on **OAuth2**.

It just extends OAuth2 specifying some things that are relatively ambiguous in OAuth2, to try to make it more interoperable.

For example, Google login uses OpenID Connect (which underneath uses OAuth2).

But Facebook login doesn't support OpenID Connect. It has its own flavor of OAuth2.