koin VS KEEP

Let's see how we can use Koin to achieve Dependency Injection (DI) and incorporate a third-party iOS Swift SDK in a KMM project. Koin supports KMM development, making it the ideal choice for KMM projects.

Might want to tell koin that : https://insert-koin.io/

New modules allow you to easily build KVision apps with the Ballast opinionated application state management framework. You can see how Ballast (together with Koin) can help you design your application architecture in the new todomvc-ballast example.

I started looking into using Kotlin/JS, and hopefully reusing a bit of code that I have, which is using Dagger and Anvil - which of course are JVM only. So I've been looking for other solutions, namely Koin and Kodein. Koin's multibinding support is... not really amazing (e.g. here, and while Kodein does support multibinding, it doesn't seem to support things like that at the declaration site - everything needs to be specified in a "DI container" (module).

Koin

It's a Koin issue, not compose. Discussed here https://github.com/InsertKoinIO/koin/issues/1079

Now I will finish up saying Hilt is just an opinionated way to dependency injection on Android, but there are simpler frameworks out there such as Koin or Kodein that can help take away a lot of the troubles that can come when you try to make your own framework. Just know that most other frameworks tend to be runtime injection instead of compile time injection like Hilt, which can be helpful to know at compile time if you are missing a dependency in your graph as opposed to runtime when its out in the wild.

I've been using Angular a lot and I like how it works. I'm also a huge fan of koin for kotlin.

for injection, I suggest Koin (https://insert-koin.io/)

`Either foo()` and `Foo foo() throws MyError` and are pretty much isomorphic.

https://github.com/Kotlin/KEEP/blob/master/proposals/stdlib/...

Nice. So for example, it looks like Kotlin has a nearly identical feature at the language level which will be optimizable when Valhalla ships: https://github.com/Kotlin/KEEP/blob/master/notes/value-class...

> In the future, in a Valhalla-capable JVM, JVM primitive classes will enable efficient representation of Kotlin value classes with an arbitrary number of underlying fields on JVM.

Most other languages agree that checked exceptions are not good by not having them.

As for alternatives, Try/Result and similar monads have decent adoption even in Java, but personally I quite like the Kotlin philosophy [1] to not have generic error containers and either use runtime exceptions or make failures of the return type.

[1] https://github.com/Kotlin/KEEP/blob/master/proposals/stdlib/...

If you want to read more and don't want to google it: https://github.com/Kotlin/KEEP/blob/data-objects/proposals/data-objects.md

Under the hood: https://github.com/Kotlin/KEEP/blob/master/proposals/coroutines.md .

you can check more here

You can use explicit API mode, then everything needs explicit visibility