Reconcile loop called every second on OpenShift, but not on Kubernetes

[jpkrohling]

When running in Kubernetes, the reconcile loop is being called only when the CR is changed, while in OpenShift it's being called every second.

https://groups.google.com/d/msg/operator-framework/DZynMacOgoY/RWaSFxFUBwAJ

#203 might be caused by this.

[jpkrohling]

This is actually caused by storing back the CR at the end of the reconciliation loop. We need to figure out a way to not update the CR in case we didn't change it, otherwise, OpenShift will trigger a new request. Kubernetes, on the other hand, seems to be smart enough and won't trigger an object update if there are no changes.