Our readers will know that inventiveness can usually be disproven (some say that the "unless" in Art. 56 EPC means that there is a legal presumption of inventiveness) but not proven. The defender of a patent or application has to rely on secondary indicia and is otherwise limited to criticise the arguments of his challenger.
This decision in very precious in that it establishes a positive test for inventive step and does this in a very clear and definitive way.
A rough translation of the an excerpt of item 5.1.3 if the decision from the French is as follows (emphasis added):
The problem-solution approach ... further leads to the rejection of any analysis according to which the person skilled in the art would have modified the closest prior art in a way which would be in conflict with its reason of existence unless its very quality as "closest prior art" originally accepted for that document would be negated.
In other words, the finding that the invention as claimed departs from the disclosure of a document of the prior art in what are - in view of the objectives at which that prior art aims - the foundations of the latter, is as such sufficient to conclude that an inventive step of the invention vis-à-vis this prior art exists.
In the case at issue, if D1 is kept as the closest prior art, the skilled person would have excluded every modification going against the aim of the assembly which is described therein, i.e. every modification which would result in a reductions of exchange within the refrigerating fluid.