I agree and disagree. FP's strength is you can layout your code in whichever way makes most sense for your project: by what they operate on or by what they do.

I agree and disagree. FP's strength is you can layout your code in whichever way makes most sense for your project: by what they operate on or by what they do. Grouping by operation makes sense until a certain point, after which you have to subdivide. If I had an FP based word processor example with 3 operations but 60 items I'd probably group by items, not operations. Imagine if FP forced every append-to-collection-like operation into a single source file. It would be madness.