Since I've written quite a bit this month, it has spurred some good conversations. I wanted to expand a bit on a question from Scott Nimrod (@Bizmonger).
As a side note, I got to meet Scott in person at NDC London back in January.
Are You Done with OO?
Here's a question Scott asked me a few days ago (Twitter link):
I've done Object-Oriented Programming for many years, and I've been successful with it. I've done Functional Programming for a much shorter time; and I really like it. But I don't see one paradigm replacing the other in my world.
These are different techniques with different strengths.
Discovering the Good
The thing I've been most impressed about in the Functional Programming community is the emphasis on discovering the good. And I've written about this before: An Outsider's View of the Functional Community.
Rather than getting caught up in "my language is better than your language", the people I know are more interested in the what each does well. When we find out what each tool is really good at, we know what to reach for when a particular problem comes up in our environment.
Even though there is an overwhelming amount of positive that I see in the functional community, every so often, I'll come across a really depressing article like "Why I'll Never Do OO Programming Again". This is often a rant about failures of projects and problems that arose in their environment.
And this makes me really sad.
Object Oriented Programming has served a good purpose over the years. And it will continue to do that.
I Really Love Object Oriented Programming
Object-Oriented programming techniques have enabled a lot of successes in my career. There are situations where having state and behavior travel together works very well.
I've programmed a lot of line-of-business applications over the years -- and having a mutable object that knows how to validate itself based on a set of rules is really great in that environment. Encapsulation is great for exposing behaviors to the outside world but hiding the implementation. When you combine this with objects that are easily data-bound to a UI, then things get even better.
Object-oriented programming is the right technique for a lot of situations.
It's also the wrong technique for others. It's difficult to multi-thread processes when there is shared, mutable state. But that's okay. When we recognize that, we can look for other techniques.
I Really Love Functional Programming
Functional programming techniques are awesome. I managed to pick up on a lot of functional-ish ideas in my C# programming with delegates, lambdas, and LINQ. I didn't realize it at the time, but these were pushing me toward a better understanding of functional programming.
By having functions with discrete inputs (parameters) and discrete outputs (return) with no shared state, things get *extremely* easy to parallelize. And I've used this techniques in certain areas to get precisely this advantage.
Functional programming is the right technique for a lot of situations.
It's also the wrong technique for others. Data-binding immutable objects doesn't make sense since data-binding is all about notification of state change. But that's okay. When we recognize that, we can look for other techniques.
There's Enough Love to Go Around
I purposely stayed away from referring directly to languages here. In the .NET world, C# is an object-oriented language, but it also has a lot of features that enable functional programming. At the same time, F# is a functional language, but it has features that enable objected oriented programming (particularly since it needs to interop with .NET libraries that are OO-focused).
Object-Oriented Programming and Functional Programming are both awesome. They just happen to be awesome at different things.We should really strive to use the right tool for the job. This means that we keep learning new things to add to our toolbox. I don't throw out my angle grinder because I got a new cordless drill. And I don't throw out perfectly working programming techniques just because I pick up a new one.
Keep learning; keep expanding; keep experimenting. What's best for your environment isn't necessarily what's best for my environment. And that's okay.