Improvements following Bridget's feedback

slides/k8s/build-with-kaniko.md

@@ -190,6 +190,8 @@ spec:
 
 - The ultimate building tool will probably be [Jessica Frazelle](https://twitter.com/jessfraz)'s [img](https://github.com/genuinetools/img) builder
 
+  (it depends on upstream changes that are not in Kubernetes 1.11.2 yet)
+
 But ... is it all about [speed](https://github.com/AkihiroSuda/buildbench/issues/1)? (No!)
 
 ---

slides/k8s/configuration.md

@@ -247,6 +247,10 @@
 
   - as holding individual configuration parameters
 
+*Note: to hold sensitive information, se can use "Secrets", which
+are another type of resource behaving very much like configmaps.
+We'll cover them just after!*
+
 ---
 
 ## Configmaps storing entire files

slides/k8s/owners-and-dependents.md

@@ -116,6 +116,26 @@ so the lines should not be indented (otherwise the indentation will insert space
 
 ---
 
+class: extra-details
+
+## When and why would we have orphans?
+
+- If we remove an owner and explicitly instruct the API to orphan dependents
+
+  (like on the previous slide)
+
+- If we change the labels on a dependent, so that it's not selected anymore
+
+  (e.g. change the `run: yanginx` in the pods of the previous example)
+
+- If a deployment tool that we're using does these things for us
+
+- If there is a serious problem within API machinery or other components
+
+  (i.e. "this should not happen")
+
+---
+
 ## Finding orphan objects
 
 - We're going to output all pods in JSON format

slides/k8s/portworx.md

@@ -92,7 +92,7 @@
 
 - For simplicity, we are going to use the latter option
 
-  (but we could use the Consul service that we deployed earlier, too)
+  (but if we have deployed Consul or etcd, we can use that, too)
 
 ---
 

slides/k8s/statefulsets.md

@@ -169,7 +169,7 @@ spec:
 
 - And arbitrary paramters for that provisioner
 
-  (replications levels, type of disk ... anything relevant!)
+  (replication levels, type of disk ... anything relevant!)
 
 - It is necessary to define a Storage Class to use [dynamic provisioning](https://kubernetes.io/docs/concepts/storage/dynamic-provisioning/)
 
@@ -371,7 +371,7 @@ nodes and encryption of gossip traffic) were removed for simplicity.
 
 - That's because we don't have a storage provider yet
 
-  (except if you're running this on your own on a full-featured cluster)
+  (except if you're running this on your own and your cluster has one)
 
 - What happens if we lose a pod?