Remove google_project.auto_create_network

[paddycarver]

Now that we have constraints/compute.skipDefaultNetworkCreation as an option, can we get rid of the auto_create_network field and logic on google_project? This would offer a more approved way to skip the default network creation.

[Chupaka]

Isn't this Organization Policy? What about standalone projects, not tied to any Org?

[morgante]

I'm a strong -1 on this change. The organization policy is set and managed on an organization level, not necessarily by the same people who are creating and updating individual projects. In many cases, my organization might not have configured the org policy but my team's projects don't want the default network.

If we did this, we'd force people to go back to the old days of writing manual bash scripts to delete the network in such cases. :(

[paddycarver]

Hey y'all, sorry for the lack of clarity here. Yes, this is talking about organization policy.

For context, I created this issue not because we have decided to do this, but because I want to remember to make an explicit decision when we're deciding what is and is not going into 3.0.0. Use cases this would cut off, like you shared, are very helpful in making that decision.

Right now I'd say probably don't worry about this, cutting off use cases with little gain except simplification is a tough sell, but I do want to talk about it with the other maintainers. I'd say going ahead with this is unlikely, but possible if we talk about it and find it's warranted or necessary.

Mostly, I just want to make an explicit decision and record it, instead of deciding by default by forgetting to decide anything. :)

[paddycarver]

So we discussed this change, and we decided that because not everyone has an org, and not everyone wants to manage this at the org-level (some users, reasonably, may want different per-project settings), we're going to keep this as-is. However, our ideal solution is for the API to create a project to support this functionality. I'm going to keep this open to track the work item of filing an internal issue to add that functionality to the API, which I think @danawillow is going to take on. When the issue has been filed, this can be closed, as we're not going to be doing it.

[danawillow]

Filed b/142969742.

[ghost]

I'm going to lock this issue because it has been closed for 30 days ⏳. This helps our maintainers find and focus on the active issues.

If you feel this issue should be reopened, we encourage creating a new issue linking back to this one for added context. If you feel I made an error 🤖 🙉 , please reach out to my human friends 👉 [email protected]. Thanks!