To achieve 3, we could set the HUGO_ENV from [context.branch-deploy.environment] to dev, but it will probably propagate faster if we use a robots.txt disallow on those domains (otherwise, the <meta> tags will take until Google next indexes each page).

AFAIK, what you propose won't work. I've had to work through a similar issue for another CNCF project with multiple versions of the docs being indexed. Based on my experiences, you'll need to change each old-version branch individually (to somehow set / config it to emit noindex, nofollow as appropriate for the branch) and have it rebuilt and redeployed.

Btw, you can't use robots.txt to prevent domains from being indexed -- see https://developers.google.com/search/docs/crawling-indexing/robots/intro:

/cc @nate-double-u

As I mentioned elsewhere, I'm OOO, but I'll be glad to help with this in the new year.

@chalin It's possible if the Netelify configs are defined for all branches in master (rather than the branches themselves) as discussed here #3628 (comment), then we might only need to update master, and then trigger a re-deploy of the older Netelify branches.

(However, I think the super new version of Hugo running in master will probably break our really old Docsy versions and the deploy might fail).

This has been done mostly as part of #3863 (comment).

We still need to update the CNAME for the older branches, but we are working on this in #3915

/close

@thesuperzapper: Closing this issue.