With `ignore_fail: true`, in the event that the upstream hasn't advanced, the fork-sync action *still takes 4 minutes* to complete

[maze88]

The sync-fork action is hardcoded to retry 4 times, with a 60 second delay. Although this is useful in certain cases, it is unnecessarily slow in other cases.

Classic example use case

• Automatically sync my fork from the upstream, on a certain schedule (e.g. nightly).
• If the upstream has new commits, everything works fine.
• But if the upstream hasn't got new commits, then the issue occurs.

Expected behavior

If the upstream hasn't got new commits, then fail fast - without retries, etc.

Suggested solution approach

Make the action idempotent - before creating the PR, add a check that there even is a diff between the Head and Base references.