Print CCR hashes in run logs

[fjahr]

This is an experiment with the suggestion from #96. This extracts the CCR hashes from rpki-client output. These hashes are potentially nicer to compare rpki cache contents because they deduplicate automatically for example.

First downside is that we need to raise the minimum rpki-client version to 9.6 because older versions don't have these CCR hashes. Second downside is that these are not being calculated in file mode in a way that is comparable afaict so we seem to need to do duplicate work in the case of reproduction runs. We can probably do this smarter and just get rid of file mode usage or maybe there is a way to get it to work in file mode, but I haven't spent enough time on this yet.

I have only done shallow testing so far. At least with the additional -P params the hashes have matched between the reproduction runs and the original downloads. Maybe that's making it worth to check the results on the latest run which has failed but I am not sure how actionable the results would be.

[jurraca]

concept ACK, it makes sense to leverage this CCR representation.

However I can't seem to get any output. Every run has empty result.stdout, so nothing is printed in output or log file. Am I holding it wrong?

regarding file mode, if you mean using the -f flag, then #97 is relevant since it removes that flag (we decided to park the change for now).

[fjahr]

However I can't seem to get any output. Every run has empty result.stdout, so nothing is printed in output or log file. Am I holding it wrong?

In nix right, I only tested without it, will investigate.

regarding file mode, if you mean using the -f flag, then #97 is relevant since it removes that flag (we decided to park the change for now).

yeah, I looked into it again but I still think that making the reproduction run a bit slower is still better than changing our decision making. I hope they can fix this omitting of the values we are using in the context of -p on the rpki-client side...

[fjahr]

However I can't seem to get any output. Every run has empty result.stdout, so nothing is printed in output or log file. Am I holding it wrong?

I think I have fixed it, can you check if it works for you now? I will still need to spend a bit more time on it and make sure this is the way to do it.

[jurraca]

ACK 43f86d7
yes, it works now. Thanks for adding the explanation.

[fjahr]

I had to fix a typo in the readme that I introduced, so I think this is mergable. I didn't find a better approach for printing those hashes and it would be good to see them in action soon.