Merge pull request crossplane#5467 from avoidalone/master

chore: remove repetitive words

design/defunct/one-pager-consuming-k8s-clusters.md

@@ -304,7 +304,7 @@ Because the scheduling of `KubernetesApplication` resources is now isolated to
 target the `KubernetesTarget` resource, more intelligent scheduling can be
 enabled without touching other parts of the Crossplane ecosystem. Previously, a
 `KubernetesCluster` claim was used for claiming, consuming, and dynamically
-provisioning Kubernetes cluster resources so changes to the the API type related
+provisioning Kubernetes cluster resources so changes to the API type related
 to scheduling (i.e. consuming) could unintentionally affect those other
 capabilities as well. Potential future scheduling improvements could involve
 price, latency, and geographic optimization by surfacing additional fields or

design/defunct/one-pager-default-resource-class.md

@@ -36,7 +36,7 @@ that can be consumed in a portable manner across cloud providers.
 
 ### Original State
 
-Originally, resource claims had to explicitly declare the the underlying
+Originally, resource claims had to explicitly declare the underlying
 resource class that they want to inherit the configuration from on deployment.
 For example, the following resource class could be declared for a Postgres RDS
 database instance on AWS:

design/defunct/one-pager-strongly-typed-class.md

@@ -86,7 +86,7 @@ spec:
   engineVersion: "9.6"
 ```
 
-This model is powerful because it allows an application developer to create a resource claim without having to know the implementation details or even the underlying provider. However, the fact that every resource class is of the same `kind` presents a key issue: The required parameters for a resource class may vary widely, and they are currently only provided as an arbitrary map that is eventually read by the controller for the the specified `provisioner`. Therefore, an administrator who is creating resource classes does not know what fields are required and will not be notified of missing or extraneous fields until the provisioning of a resource that references the class.
+This model is powerful because it allows an application developer to create a resource claim without having to know the implementation details or even the underlying provider. However, the fact that every resource class is of the same `kind` presents a key issue: The required parameters for a resource class may vary widely, and they are currently only provided as an arbitrary map that is eventually read by the controller for the specified `provisioner`. Therefore, an administrator who is creating resource classes does not know what fields are required and will not be notified of missing or extraneous fields until the provisioning of a resource that references the class.
 
 The `parameters` supplied by the resource class are used to populate the `spec` of the managed resource (i.e. the Kubernetes representation of the external resource) when it is created. For instance, the creation of `mysql-claim`, which references the `standard-mysql` class, is watched by the claim controller for AWS RDS instances. It brings together the information provided in the claim and class to create the `RDSInstance` managed resource. Specifically, it calls the `ConfigureMyRDSInstance()` function. As part of the configuration, the function creates the `spec` of the `RDSInstance` managed resource from the `parameters` of the `ResourceClass`:
 

design/design-doc-packages-v2.md

@@ -161,7 +161,7 @@ workflow the package manager uses for installing a `Package`. We will use a
    [stdout](https://github.com/crossplane/crossplane/blob/a0d139f7cf269599ba916ed15af3fd68ffeabbdf/cmd/crossplane/package/unpack/unpack.go#L53).
 4. The _`ClusterPackageInstall` controller_ waits for the `Job` to complete
    successfully before reading the logs from the `Pod`. When the `Job` is
-   complete, it reads the the logs and creates all of the objects that were
+   complete, it reads the logs and creates all of the objects that were
    printed, making a [few modifications as well as annotating and labelling
    appropriately](https://github.com/crossplane/crossplane/blob/6fc50822fbf11a7d31f8a9dabde5c8948c3b36ac/pkg/controller/packages/install/installjob.go#L259).
 5. The _`Package` controller_ observes the `Package` creation and assumes the
@@ -714,7 +714,7 @@ will operate with the following behavior.
    `spec.desiredState: Active`.
 3. Set `status.currentRevision` to full image name used for `PackageRevision`
    (this can be obtained from the `Pod` in the install `Job`)
-4. Report status of `PackageRevision` in the the status of the `Package`.
+4. Report status of `PackageRevision` in the status of the `Package`.
 
 **`Package` Created with `spec.revisionActivationPolicy: Manual`**
 
@@ -723,7 +723,7 @@ will operate with the following behavior.
    `spec.desiredState: Inactive`.
 3. Set `status.currentRevision` to full image name used for `PackageRevision`
    (this can be obtained from the `Pod` in the install `Job`)
-4. Report status of `PackageRevision` in the the status of the `Package`.
+4. Report status of `PackageRevision` in the status of the `Package`.
 
 User is responsible for manually setting the `PackageRevision` to `Active`.
 

design/one-pager-resource-connectivity-mvp.md

@@ -688,7 +688,7 @@ resources need to be created beforehand:
   and associating it with a set of subnets.
 
 In addition, `RDSInstance`s also need the following resources, so that they are
-accessible by the the worker nodes:
+accessible by the worker nodes:
 
 * `DBSubnetGroup`: represents a group of `Subnet`s from different availability
   zones,
@@ -964,7 +964,7 @@ ensure connectivity.
 
 We will need to wait until *after* the Wordspress stack is installed to create
 the VNet Rule on the MySQL DB due to the fact that the database will not exist
-until the the stack references our `SQLServerClass` with a claim.
+until the stack references our `SQLServerClass` with a claim.
 
 #### A Model for Deploying Wordpress