119
How much are SOLID principles in OOP programming (and JVM languages specifically) just a mindless following of a set of ideas that aren't always the best solution? (programming.dev)
submitted 1 year ago by to c/programming@programming.dev
110 comments fedilink hide all child comments
 

As a Java engineer in the web development industry for several years now, having heard multiple times that X is good because of SOLID principles or Y is bad because it breaks SOLID principles, and having to memorize the "good" ways to do everything before an interview etc, I find it harder and harder to do when I really start to dive into the real reason I'm doing something in a particular way.

One example is creating an interface for every goddamn class I make because of "loose coupling" when in reality none of these classes are ever going to have an alternative implementation.

Also the more I get into languages like Rust, the more these doubts are increasing and leading me to believe that most of it is just dogma that has gone far beyond its initial motivations and goals and is now just a mindless OOP circlejerk.

There are definitely occasions when these principles do make sense, especially in an OOP environment, and they can also make some design patterns really satisfying and easy.

What are your opinions on this?

you are viewing a single comment's thread
view the rest of the comments
[–] 14 points 1 year ago (7 children)

The main thing you are missing is that "loose coupling" does not mean "create an interface". You can have all concrete classes and loose coupling or all classes with interfaces and strong coupling. Coupling is not about your choice of implementation, but about which part does what.

If an interface simplifies your code, then use interfaces, if it doesn't, don't. The dogma of "use an interface everywhere" comes from people who saw good developers use interfaces to reduce coupling, while not understanding the context in which it was used, and then just thought "hey so interfaces reduce coupling I guess? Let's mandate using it everywhere!", which results in using interfaces where they aren't needed, while not actually reducing coupling necessarily.

  • permalink
  • fedilink
  • source
  • hideshow 7 child comments
    •
  • [–] 6 points 1 year ago (5 children)

    I think a large part of interfaces everywhere comes from unit testing and class composition. I had to create an interface for a Time class because I needed to test for cases around midnight. It would be nice if testing frameworks allowed you to mock concrete classes (maybe you can? I haven't looked into it honestly) it could reduce the number of unnecessary interfaces.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 5 child comments
    •
  • [–] 4 points 1 year ago (1 child)

    At least in C# with Moq you can only mock virtual methods of concrete classes, so using interfaces is still nicer in general.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 1 child comment
    •
  • [–] 4 points 1 year ago

    Yeah Moq is what I used when I worked with .NET.

    On an unrelated note; god I miss .NET so much. Fuck Microsoft and all that, but man C# and .NET feels so good for enterprise stuff compared to everything else I've worked with.

  • permalink
  • fedilink
  • source
  • parent
  • [–] 2 points 1 year ago

    I've had to do that too, for tests specifically as well, and making clocks an interface on the spot was trivial. I did it when I needed it though, and not ahead of time.

    A Time interface is waaaay too broad. Turns out, I only needed something something to give me programmable ticks for my tests, which is much narrower in scope than abstracting something as general as time.

    I'd say abstractions designed to support tests need to be very narrow in scope, and focused on solving the problem at hand.

  • permalink
  • fedilink
  • source
  • parent
  • [–] 3 points 1 year ago

    As a dev working on a large project using gradle, a lot of the time interfaces are useful as a means to avoid circular dependencies while breaking things up into modules. It can also really boost build times if modules don't have to depend on concrete impls, which can kill the parallelization of the build. But I don't create interfaces for literally everything, only if a type is likely going to be used across module boundaries. Which is a roundabout way of saying they reduce coupling, but just noting it as a practical example of the utility you gain.

  • permalink
  • fedilink
  • source
  • parent