https://docs.google.com/document/d/1iZx-In4pnq5Op53V8B2ipGWmQCiSK1hT8nRH3NL6FdE/edit#heading=h.58ek1br972wd

Hashicorp’s move to adopt the Business Source License has inevitably led to discussions regarding how we move off Terraform and what the plan would be for us doing that. Ansible has been raised as a potential replacement but given that we have been there in openshift-3x we likely don’t want to go back down that road. CAPI potentially provides a very viable solution and one that has already been proposed as the direction we should be heading prior to the Hashicorp announcement. With the current development point of CAPI, however, that raises several questions around how that aids our goals to get to a simplified Installation experience with fewer installers, and the need to go back to the basics of what are the actual requirements and base needs for our customers.

Key requirements and end goals of whatever solution we provide

The approach should lead to a simplified experience with fewer installers - simplify the decision tree to get to OpenShift
We really need to focus on the actual requirements of our important customers and how they use the installers in practice. Demonstrate that we’re going to address the real needs.