Still incomplete.

Passing no value argument to the argument makes podman look it up in
the current process, which is exactly what we want here.

Just the template and some notes on installation for now.

See also team#40972 and tpo/tpa/team#40929...

Closes: team#40478

They were handled without noted problems in the bullseye first batch,
I don't see why change it here.

This was done by running this on pauli. First list buster machines:

  580  curl -s -G http://localhost:8080/pdb/query/v4 --data-urlencode 'query=nodes { facts { name = "lsbdistcodename" and value = "buster" }}' | jq -r .[].certname | sort

Then list bookworm machines:

  581  curl -s -G http://localhost:8080/pdb/query/v4 --data-urlencode 'query=nodes { facts { name = "lsbdistcodename" and value = "bookworm" }}' | jq -r .[].certname | sort

Then list the other machines, presumably bullseye:

  582  curl -s -G http://localhost:8080/pdb/query/v4 --data-urlencode 'query=nodes { facts { name = "lsbdistcodename" and value != "bookworm" and value != "buster" }}' | jq -r .[].certname | sort

Check that the count matches (i.e. that there is no other release pending):

  583  curl -s -G http://localhost:8080/pdb/query/v4 --data-urlencode 'query=nodes { facts { name = "lsbdistcodename" and value != "bookworm" and value != "buster" }}' | jq -r .[].certname | sort | wc -l
  584  curl -s -G http://localhost:8080/pdb/query/v4 --data-urlencode 'query=nodes { facts { name = "lsbdistcodename" and value = "bullseye" }}' | jq -r .[].certname | sort | wc -l

List the bullseye machines:

  585  curl -s -G http://localhost:8080/pdb/query/v4 --data-urlencode 'query=nodes { facts { name = "lsbdistcodename" and value = "bullseye" }}' | jq -r .[].certname | sort

Then the latter list was pasted in the first batch, and manually split
out.

We specifically address the Python 2 retirement here, so i think this
deals with team#40999 as well, so:

Closes: team#40999

Turns out you just can't have pipes in a markdown table... ?

I thought it would export only subkeys and therefore that
export-secret-keys would not, but it's more complicated than that:
export-secret-subkeys exports all secret keys (including the primary)
but in a way that "render the secret part of the primary key useless".