Let me turn your attention to the comparison of memory consumption of the previously mentioned object-oriented and functional solutions. The object-oriented solution eagerly creates, materializes in computer memory the collection of 996 OfDigits objects; that is, its memory consumption is a linear function of the problem dimensions. In contrast to this approach, the functional solution does not require more than a single instance of OfDigits at any moment of the max aggregation, lazily producing the same 996 objects one by one according to the demand of the max aggregator function, hence having memory consumption that is constant and (almost) independent of the problem dimensions. This is a rather complex quality. If you imagine that the initial condition has suddenly changed and hugeNumber is really huge, then the object-oriented solution may become non-applicable due to the lack of required memory, while the functional solution, being agnostic to this factor, will continue to work. Generalizing this observation, the functional paradigm allows you to solve problems of a bigger scale, rather than taking other approaches by utilizing the lazy manner of data manipulation. The interesting corollary stemming from this approach is the technique of manipulating data sequences of unlimited length that do not require their complete materialization in the memory.