Confused about the Interface and Class coding guidelines for TypeScript

I read through the TypeScript Coding guidelines

And I found this statement rather puzzling:

Do not use "I" as a prefix for interface names

I mean something like this wouldn't make a lot of sense without the "I" prefix

class Engine implements IEngine

Am I missing something obvious?

Another thing I didn't quite understand was this:

Classes

For consistency, do not use classes in the core compiler pipeline. Use function closures instead.

Does that state that I shouldn't use classes at all?

Hope someone can clear it up for me :)

When a team/company ships a framework/compiler/tool-set they already have some experience, set of best practices. They share it as guidelines. Guidelines are recommendations. If you don't like any you can disregard them. Compiler still will compile your code. Though when in Rome...

This is my vision why TypeScript team recommends not I-prefixing interfaces.

Reason #1 The times of the Hungarian notation have passed

Main argument from I-prefix-for-interface supporters is that prefixing is helpful for immediately grokking (peeking) whether type is an interface. Statement that prefix is helpful for immediately grokking (peeking) is an appeal to Hungarian notation. I prefix for interface name, C for class, A for abstract class, s for string variable, c for const variable, i for integer variable. I agree that such name decoration can provide you type information without hovering mouse over identifier or navigating to type definition via a hot-key. This tiny benefit is outweighed by Hungarian notation disadvantages and other reasons mentioned below. Hungarian notation is not used in contemporary frameworks. C# has I prefix (and this the only prefix in C#) for interfaces due to historical reasons (COM). In retrospect one of .NET architects (Brad Abrams) thinks it would have been better not using I prefix. TypeScript is COM-legacy-free thereby it has no I-prefix-for-interface rule.

Reason #2 I-prefix violates encapsulation principle

Let's assume you get some black-box. You get some type reference that allows you to interact with that box. You should not care if it is an interface or a class. You just use its interface part. Demanding to know what is it (interface, specific implementation or abstract class) is a violation of encapsulation.

Example: let's assume you need to fix API Design Myth: Interface as Contract in your code e.g. delete ICar interface and use Car base-class instead. Then you need to perform such replacement in all consumers. I-prefix leads to implicit dependency of consumers on black-box implementation details.

Reason #3 Protection from bad naming

Developers are lazy to think properly about names. Naming is one of the Two Hard Things in Computer Science. When a developer needs to extract an interface it is easy to just add the letter I to the class name and you get an interface name. Disallowing I prefix for interfaces forces developers to strain their brains to choose appropriate names for interfaces. Chosen names should be different not only in prefix but emphasize intent difference.

Abstraction case: you should not not define an ICar interface and an associated Car class. Car is an abstraction and it should be the one used for the contract. Implementations should have descriptive, distinctive names e.g. SportsCar, SuvCar, HollowCar.

Good example: WpfeServerAutosuggestManager implements AutosuggestManager, FileBasedAutosuggestManager implements AutosuggestManager.

Bad example: AutosuggestManager implements IAutosuggestManager.

Reason #4 Properly chosen names vaccinate you against API Design Myth: Interface as Contract.

In my practice, I met a lot of people that thoughtlessly duplicated interface part of a class in a separate interface having Car implements ICar naming scheme. Duplicating interface part of a class in separate interface type does not magically convert it into abstraction. You will still get concrete implementation but with duplicated interface part. If your abstraction is not so good, duplicating interface part will not improve it anyhow. Extracting abstraction is hard work.

NOTE: In TS you don't need separate interface for mocking classes or overloading functionality. Instead of creating a separate interface that describes public members of a class you can use TypeScript utility types. E.g. Required<T> constructs a type consisting of all public members of type T.

export class SecurityPrincipalStub implements Required<SecurityPrincipal> {
  public isFeatureEnabled(entitlement: Entitlement): boolean {
      return true;
  }
  public isWidgetEnabled(kind: string): boolean {
      return true;
  }

  public areAdminToolsEnabled(): boolean {
      return true;
  }
}

If you want to construct a type excluding some public members then you can use combination of Omit and Exclude.