Why is namespacing implemented completely differently to Vuex?

[geoidesic]

So in Vuex the way you set up namespacing is like so:

let store = new Vuex.Store({
    modules: {
          moduleA: {
               getters:{}
          }
    }
})

According to the docs for this repo, it sets up namespaces store in a completely different fashion:

new Store({
  state: {
    value: 'from root',
    moduleA: {
      value: 'from A',
      moduleC: {
        value: 'from A/C',
      },
    },
    moduleB: {
      value: 'from B',
    },
  },
})

This is counter-intuitive and makes it impossible to automatically create a store. I'm still having to manually mock everything if I have to completely rewrite it. So I'm struggling to understand what the point of this library is?

Sorry, not meant to be a rant :) just trying to understand.

Maybe I have the complete wrong end of the stick, but my thinking is that I have a namespaces set of stores already. I'm happy with them as is. I just want them mocked, so I can run tests. I really, really don't want to have to manually map every getter, mutation and state pair in order to achieve a mocked store and I can't imagine what the purpose of this library is if it's not automating that?

[posva]

Yes, it's different. Not having to declare mutations nor actions allows to have a more straightforward way with one single state object which is the result of having modules instead of having to write one state and as many modules as needed.

About reusing existing store stuff: #9