Indicate splits which lost PRs somehow #1216
Description
From time to time, people notice that a failing has failed due to a specific PR, and r- that PR. Currently this is quite difficult to suss out:
- there is no explicit link between a staging and its splits so only the top-level link is easy to make
- there is no indication or counter, so knowing if a PR (really batch) disappeared between the original failing and the split being staged is difficult
- which makes realising that a PR has been r-'d hard to figure out if not told, you basically have to hunt down the culprit as well, and find that it's been unreviewd 15mn earlier