I don't find it controversial, and it isn't altogether interesting, but I would like it if it were better communicated. I've cleaned up code that dogmatically followed a interface-per-class policy.

It's quite controversial in the real world where "each service class has an interface" is official codified dogma.

OTOH in certain specific cases I see the value of the interface even for just one class when I want to clearly separate the contract (which I'm bound to fulfill now and in the future) and the implementation specific comments.

Let's say that in my little pet programming language I have just one List type. I want to clearly separate what's the contract (indexed ordered sequence) as opposed to "incidental" implementation details (constant cost random index access).

I have worked at shops where, without fail,

  EveryClass : IEveryClass

Because, of course, you might expand it later.

Sure, I'm adding this shitty console based logger, but since it's wrapped in ILogger I'll later come and write a FileBasedLogger, a DBLogger and a KafkaProducerLogger for when we go webscale.

Except, that never happens.

  //Names changed slightly to protect the wicked
  public interface IProvideBoolean {
    Boolean True { get; }
    Boolean False { get; }
  }

  public class BooleanProvider : IProvideBoolean {
    public Boolean True => true;
    public Boolean False => false;
  }

So.... SOLID.... If there's ever a third value for Boolean, I Will Be Ready.

See also: https://thedailywtf.com/articles/What_Is_Truth_0x3f_

Finally ready for trinary computing!

Then you go back and refactor the code to pull out the interface! I so hate interface abuse.

If you use names like Logger instead of ILogger, you don't even have to change all the names when you refactor to an interface! If you end up wanting a FileBasedLogger, you just rename the current implementation to StdoutLogger, create a Logger interface, create a FileBasedLogger, and change instantiation sites to build the instance you want.

> If you use names like Logger instead of ILogger, you don't even have to change all the names when you refactor to an interface!

But you want to, that's the point. You're making a drastic change to your code base, surely you want to inspect every place where that type is used.

Using IClass naming is considered best practice in the .Net world AFAIK. (I think Visual Studio and/or Resharper hints very clearly about this.)

> Using IClass naming is considered best practice in the .Net world AFAIK.

No, it's the opposite. IClass is the norm, for legacy reasons going all the way back to the early COM days (it all started with IUnknown[1]).

[1] https://msdn.microsoft.com/en-us/library/windows/desktop/ms6...

I know, I'm demonstrating one reason I think that's a bad best practice :) I recognize there's nothing most people can do about that, though!

There are some (rare) occasions, when private members drag new dependencies that you don't want the user of your class to know about.

You can also do "pImpl", which is similar, if not worse, in terms of boilerplate, but doesn't allow DI.

Just don't use these techniques unless you can prove they're needed (e.g private members drag new problematic dependencies, #include <some_volatile_header_file.h>).

The rule of three works out pretty well. If you don't have three of something then you don't really have a group you can attribute a set of behaviors to. If and when a third one does present itself, you usually have to rework things because you were very wrong about the way you organized things.

> That's not really controversial or interesting.

It is to me. So far it seems that everyone takes this for granted. I disagree: I have yet to see a problem created by the "each class has an interface" rule, and I have seen countless problems created by classes without one.

So: what problems are caused by interfaces with only one implementation?