CI: check apidiff once v1beta1 is released

[EmilienM]

/kind feature

Describe the solution you'd like
Once we release CAPO v1beta1, our API should remain stable.

One way to check that in CI is to use github.com/joelanford/go-apidiff as it's already done in other providers.
We could add a job that will run apidiff against PR and return an error if the API change is major and not minor.

[mdbooth]

This is a great idea in principal. I'm not convinced this specific library will help us, though, as it seems to be checking Go semantics rather than K8S semantics, and could therefore give us both false-positives (e.g. altering a Go field in a manner which doesn't affect the CRD) and false-negatives (e.g. altering a kubebuilder marker in a comment which is not visible in the Go semantics, but does alter the CRD).

That said, in light of #1891 we may want to start caring about the former, which we have not done previously.

[EmilienM]

Azure has it:
https://prow.k8s.io/job-history/gs/kubernetes-jenkins/pr-logs/directory/pull-cluster-api-provider-azure-apidiff

And the way they use it is pretty smart:
https://storage.googleapis.com/kubernetes-jenkins/pr-logs/pull/kubernetes-sigs_cluster-api-provider-azure/4585/pull-cluster-api-provider-azure-apidiff/1760447267543715840/build-log.txt

+ WRAPPED_COMMAND_PID=18
+ wait 18
+ ./scripts/ci-apidiff.sh
*** Running go-apidiff ***
No files under api/ or exp/api/ changed.
+ EXIT_VALUE=0
+ set +o xtrace

So I really propose that we find a way to have something similar.

[k8s-triage-robot]

The Kubernetes project currently lacks enough contributors to adequately respond to all issues.

This bot triages un-triaged issues according to the following rules:

• After 90d of inactivity, lifecycle/stale is applied
• After 30d of inactivity since lifecycle/stale was applied, lifecycle/rotten is applied
• After 30d of inactivity since lifecycle/rotten was applied, the issue is closed

You can:

• Mark this issue as fresh with /remove-lifecycle stale
• Close this issue with /close
• Offer to help out with Issue Triage

Please send feedback to sig-contributor-experience at kubernetes/community.

/lifecycle stale

[EmilienM]

/remove-lifecycle stale
/assign EmilienM

This one is close to my heart, I'll take care of it.

[lentzi90]

I think we should care about the Go API also, so this would be useful.

[mdbooth]

It would be good to agree the scope of our Go stability guarantees.

We should probably also switch to proper semver for CAPO releases to account for breaking changes.