We describe a solution, written using the Whole Platform, to a selection of benchmark problems proposed to evaluate and compare language workbenches. More specifically, we selected the problems best suited to highlight the composability feature of the Whole Platform. We think that the programming languages and consequently the end user applications are mainly designed to be standalone. Composition of program functionalities is supported only in some very limited and anticipated ways, and this severely restrict the availability of the knowledge embedded in a software product. This may seem natural and unavoidable, but it is not if you take as a reference the human mind. We believe that the outcome of software development should be additional knowledge incorporated in the language workbench and that every form of modularity, including applications, should be relegated as a mean to package and distribute knowledge from a system to another.