|As an OpenShift team, we are hoping to clarify future direction with regards to big architectural changes to OpenShift. Lots of discussions about various features go on in lots of formats, and eventually those discussions need to be captured in one place. In addition, when developers come to an agreement about something, we want to make sure that we describe that agreement and why it happened.|
One model that we happen to particularly like is the open Python PEP model, where changes to the core specification are discussed in email, debated, justified, endlessly rehashed, and then finally become a proposal for a change with a concrete definition. Once the change is implemented, the PEP becomes part of the living record of the evolution of the language. For OpenShift we'd like to capture some of the large intentional changes that will happen at some point in the future, even if we don't have all the details worked out or a schedule for when it will be implemented. Knowing about what is coming -- and the rationale behind it -- helps everyone better understand where is OpenShift is and where it is going.
The public OpenShift PEP repository is located here:
The README describes the rough structure of a PEP and how we intend to track those items. As you can see, the repo is currently sparse - with only a single draft PEP related to the administrative console. We intend to translate some existing documents into PEP form and add them soon. If the process described sounds a bit rough, that's because it is. The process will evolve to need the needs of the project rather than vice versa.
It's important to note that PEPs aren't feature requests - those can be done via the public ideas page athttps://openshift.redhat.com/community/ideas. Instead, they're the congealed form of a technical consensus about a particular feature. Like Python, the correct place to suggest, propose, discuss, argue over, and eventually reach agreement on new large technical features in OpenShift is this mailing list - dev lists openshift redhat com. We hope to broaden that discussion over the coming months.
Clayton Coleman - OpenShift