FP & Network effect
Few months ago (my writing, and my publishing queue is a bit long), I've watched a video of Dave Farley about Functional Programming adoption failure.
Note: we could discuss the fact that "FP hasn't taken over the world", given that all mainstream programming languages have introduced FP-style constructions and many concepts in FP-related fields in their standard libraries.
Note 2: we could point-out that "taking over the world", hasn't been a goal of "FP" programming languages, unlike providing a comprehensive and coherent reasoning framework.
He pointed out few elements:
- Immutability doesn't improve performances of non-parallelizable problems
- Some problems are best expressed in OOP
- FP tends to push concepts too far to be usable on "mainstream project"
I let you go through the video as it covers a lot of interesting historical details.
I could spend an entire series of logs disagreeing about each element, or there relevance, but, to keep it short, I don't think the idea of FP isn't structurally flawed, and some of the current implementations are good enough.
Instead, I would focus on what I think make FP programming languages not widely adopted: network effect.
In economics, [..] is the phenomenon by which the value or utility a user derives from a good or service depends on the number of users of compatible products.
Applied to programming languages: the more users of a programming language, the more it becomes usable. You get:
- More libraries
- More Stack Overflow/Discourse/Discord/Reddit/LLMs support
- More tooling/Higher completion quality
- More blogs/books/example/tutorials/videos
- More boot-camps/trainings/school curriculums
- More people to hire
As a quick reminder, not so long ago, Linux and PostGreSQL had a tiny market share compared to their competitors. They didn't radically change since then, the context changed, and they became relevant.
Maybe this time won't come for FP, but that doesn't imply FP is wrong, or shouldn't be used