Joel Speed

From your suggestions, I'd be interested to see if we can move towards Option A, though I don't know how simple that is for markers where we currently don't have...

> While rewriting the PR to have a centralised utility for supporting feature gates, I found it would require modifying each validation marker definition and the schema generation logic -...

IIUC, the ELB SG name is generated based on a hash of the name of the service, therefore, shouldn't we be able to check for the existence of a SG...

Cluster owned tag seems like a reasonably good assumption, but wouldn't we also check the name since that is predictable?

Upstream documentation already notes that this value should not be changed (see [here](https://kubernetes.io/docs/reference/labels-annotations-taints/#service-beta-kubernetes-io-aws-load-balancer-type:~:text=the%20Service%20specification.-,Caution%3A,object.%20See%20the%20AWS%20documentation%20on%20this%20topic%20for%20more%20details.,-service.beta.kubernetes)) ALBO documentation explains this as well (see [here](https://kubernetes-sigs.github.io/aws-load-balancer-controller/v2.2/guide/service/annotations/#:~:text=This%20annotation%20should%20not%20be%20modified%20after%20service%20creation.)) But there's no enforcement of this since there is...

> but would increase controller's API calls when syncronizing LBs Yes, I suspect this is not the way we would want to go because of those additional API calls that...

@oduludo please check my previous comments, doesn't look like they all got applied correctly

> Introduce a +kubebuilder:feature-gate= marker to allow Go struct fields, RBAC, and webhook configurations to be conditionally included or excluded in the generated output. Does `` here refer to one...

> @JoelSpeed I added the field names in the error messages so now it's better scoped. Let me know if I can go on doing the same for core and...

/lgtm /labe tide/merge-method-squash