the plan is to put together an integration test
of diff application to opaque data through the TreeMutator,
using the now roughly finished binding primitives.
moreover, the idea is to apply precisely the same diff sequence,
as was used in the detail test (TreeMutatorBinding_test).
NOTE: right now, the existing placehoder code applies this sequence
onto a Rec<GenNode>. This should work already -- and it does,
BUT the result of the third step is wrong. Really have to
investigate this accidental finding, because this highlights
a conceptual mismatch in the handling of mixed scopes.
the plan is to use this specific diff sequence
both in the individual binding tests, and in a
more high level integration test. Hopefully this
helps to make these quite technical tests more readable
how can ordinary object fields be treated as "Attributes"
and thus tied into the Diff framework defined thus far.
This turns out to be really tricky, even questionable
