ReactiveViewModel is a combination code/documentation project for building Cocoa applications using Model-View-ViewModel and ReactiveCocoa.
By explaining rationale, documenting best practices, and providing reusable library components, we want to make MVVM in Objective-C appealing and easy.
Most Cocoa developers are familiar with the Model-View-Controller (MVC) pattern:
Model-View-ViewModel (MVVM) is another architectural paradigm for GUI applications:
Although it seems similar to MVC (except with a "view model" object in place of the controller), there's one major difference — the view owns the view model. Unlike a controller, a view model has no knowledge of the specific view that's using it.
This seemingly minor change offers huge benefits:
- View models are testable. Since they don't need a view to do their work, presentation behavior can be tested without any UI automation or stubbing.
- View models can be used like models. If desired, view models can be copied or serialized just like a domain model. This can be used to quickly implement UI restoration and similar behaviors.
- View models are (mostly) platform-agnostic. Since the actual UI code lives in the view, well-designed view models can be used on the iPhone, iPad, and Mac, with only minor tweaking for each platform.
- Views and view controllers are simpler. Once the important logic is moved elsewhere, views and VCs become dumb UI objects. This makes them easier to understand and redesign.
In short, replacing MVC with MVVM can lead to more versatile and rigorous UI code.
A view model is like an adapter for the model that makes it suitable for presentation. The view model is also where presentation behavior goes.
For example, a view model might handle:
- Kicking off network or database requests
- Determining when information should be hidden or shown
- Date and number formatting
- Localization
However, the view model is not responsible for actually presenting information or handling input — that's the sole domain of the view layer. When the view model needs to communicate something to the view, it does so through a system of data binding.
OS X and iOS both have view (or window) controllers, which may be confusing at first glance, since MVVM only refers to a view.
But upon closer inspection, it becomes apparent that view controllers are actually just part of the view layer, since they handle things like:
- Layout
- Animations
- Device rotation
- View and window transitions
- Presenting loaded UI
So, "the view" actually means the view layer, which includes view controllers. There's no need to have a view and a view controller for the same section of the screen, though — just pick whichever class is easier for the use case.
No matter whether you decide to use a view or a view controller, you'll still have a view model.
MVVM is most successful with a powerful system of data binding. ReactiveCocoa is one such system.
By modeling changes as signals, the view model can communicate to the view without actually needing to know that it exists (similarly for model → view model communication). This decoupling is why view models can be tested without a view in place — the test simply needs to connect to the VM's signals and verify that the behavior is correct.
ReactiveCocoa also includes other conveniences that are hugely beneficial for MVVM, like commands, and built-in bindings for AppKit and UIKit.
Model-View-ViewModel was originally developed by Microsoft, so many of the examples are specific to WPF or Silverlight, but there are still a few resources that may be useful:
Blog posts:
Presentations: