Update messaging, it is used in the backend also

[paultyng]

Related to hashicorp/terraform-provider-aws#13057, its unclear, since this is a common depenedency, whether this is for a provider or backend, when it could be both. Modifying the messaging here to make that a little more obvious. Ideally we should just make downstream consumers wrap these errors.

[bflad]

Related issues:

• Remove AWS Provider verbiage from error messaging #24

I think we should completely remove mentions of which upstream caller might be involved (since it leaves it ambiguous to the operator to figure out which is applicable), instead allowing the S3 Backend/AWS Provider to return the proper error messaging context themselves.

e.g. In Terraform S3 Backend:

sess, err := awsbase.GetSession(cfg)

if err != nil {
  return fmt.Errorf("error configuring S3 Backend: %w", err)
}

e.g. in Terraform AWS Provider:

sess, accountID, partition, err := awsbase.GetSessionWithAccountIDAndPartition(awsbaseConfig)

if err != nil {
	return nil, fmt.Errorf("error configuring AWS Provider: %w", err)
}

[bflad]

Regarding the error message with the URL, that should be promoted to an exported type so the upstream code can catch that specific case and create their own appropriate messaging if they desire.

[bflad]

Going to pull this in as-is for now and iterate on it (likely along with #25). 👍 Thanks, @paultyng!