[Haskell-cafe] Engineering Value of Functional Programming

[jo at durchholz.org]

On 07.12.24 13:45, Mostafa Touny via Haskell-Cafe wrote:
> Software engineers usually deviate away from Haskell, in the name of rapid development.
> 
> In Pure Math, I can see the power of abstraction; It resulted in broad applications, with a sustainable and scalable usage in all humanity's sciences. Software development should benefit as well, avoiding technical debts and refactoring costs. Haskell seems more promising as it is empowered by category and type theory.
> 
> Nonetheless, I cannot find a single management methodology, like Eric Ries' lean startup and iterative agile, that demonstrates the power of functional programming from the perspective of project management.
> 
> Discussion.
> - Do you agree category and type theory could reduce projects costs?
> - Is it true, no guideline is designed for demonstrating their worthiness?

Frame challenge:

Most programming is application programming, and application programming 
is finding, integrating and using the best libraries for the project.

Guidelines are for application programming, they don't help much with 
libraries as each library is one of its kind (sort-of, usually there's 
one go-to libraries and maybe two, sometimes three competitors).

So I think it's normal to expect that there's nothing that teaches type 
theory or category theory with the aim of reducing project cost.
However, a lot of arguments can be made that these things make it easier 
to create good libraries.

There is one exception: DDD. A domain type is straightforward to model 
in a pure language, and often pretty boilerplatey elsewhere.
However, coding these types does not usually take a lot of time overall, 
so it's not a strong argument; having less boilerplate overall, though, 
would make a pretty strong argument, but again, that's not methodology, 
just practical experience, since the type system are given anyway so 
there's no methodology to teach.

Just my 2c.

Regards,
Jo