I'm having some fun trying to get my head around some MVP stuf, as it pertains to User Controls. I'm using .NET WinForms (or something close to it) and Supervising Controller pattern (well, I think I am :).
The User Control is itself part of an MVP application (its the View and has an associated Presenter etc). The Presenter is always started first, and it starts the Model(s) and then View(s). The View builds its UI, part of which will be to NEW the UC, which is the View.
Now the (form) Presenter needs to know about the UC Presenter, but I'm thinking that it doesn't know anything about how the View is composed. The form Presenter doesn't, for instance, know that the UC is part of the form's Controls collection, nor should it.
Furthermore, the design experience should not be changed; IOW the dev of the View (form) should just be able to select a User Control from the toolbox and drop it on a form.
So, on to my questions. Firstly, are my assumptions above correct? Somewhat misguided? Messed up? WTF are you thinking?
Secondly, is it right (enough?) to have the form View invoke the UC View, and the form Presenter invoke the UC Presenter and have some mechanism to tell the UC View what its Presenter is? This breaks my "Presenter first" rule, but I'm not sure how else to do it.
Any other thoughts, suggestions, comments gladly accepted.
-- nwahmaet
A presenter should be thought of as "autonomous state" in the presentation tier. This means that it is responsible for ensuring that the view's presentation of the model's state is in sync. The reason I bring this up is because the "pattern" of MVP often gets lost in the dogmatic view of how things should be separated. It seems that this is one reason Martin Fowler decided to try to clarify the terminology around the MVP pattern.
My favored flavor of MVP is the passive view, so my answer is based off of that.
I implement composite user controls and forms very often using the passive view pattern. There are essentially 3 different configurations:
One presenter for all user controls in the hierarchy. Flatten the view using an interface.
One presenter for each user control in the composite tree. Each parent presenter is responsible for instantiating and initializing its child presenters. The user controls are created at design time, and are able to function without a presenter (with no presentation behavior)
One presenter for each user control in the composite tree. All of the presenters are loosely coupled through a higher level controller class. The controller class is responsible for construcing the presenter, wiring them up, and coordinating their events.
Although it is a solution of last resort for me (because of its complexity), I think that the last option is the solution that you are looking for.
I've been running up against this exact problem for several months in an application I'm working on. The conclusion that I've very recently come to is that in many cases it might be impossible to apply the MVP pattern at both the window AND user control levels, without "breaking" the pattern.
My thought on it is that the user control is part of the view implementation, and the presenter should not know what is going on inside the view implementation, which means that the window-level presenter by extension should not know about the user control's presenter, and hence there should be no communication between them, including instantiation of the latter by the former. It might be argued that the user control's presenter is part of the window view implementation, and so the window view may instantiate the user control presenter. But it cannot inject the model classes that the presenter needs, because the view isn't supposed to be aware of them.
The conclusion that I think I am arriving at is that ALL user controls are view-implementation-specific, and so should be contained completely within the view silo of the larger pattern. As such, they don't get to have their own presenters... At least not bundled up with the control implementation itself. Instead they should be manipulated indirectly by the parent window's presenter, via pass-through fields exposed on the view interface. In short, the user control is exposed to the presenter not by its own interface, but rather via a common pass-through interface implemented by its parent view. Call this a "partial view interface".
Your presenter can then contain instances of a re-usable sub-presenter class which works only with this partial view interface, and the relevant pieces of the model. This will allow you to avoid re-writing the presenter code to translate from the model every time you need to use the control, AND it prevents the window view from needing to know about the model in order to pass info through to the control's presenter.
What this effectively does is it further separates the user control, as a module, from your data model. This makes sense if you think of a user control, as a whole, as an element of the view implementation. As a re-usable unit, it is a piece of view functionality, and no part of it should be tied to your data model.
Your questions is general that a variety of schemes could apply.
In this case my guess is that you should look at Observer Pattern.
You have a interface that anything that uses that view would implement. Then it would register itself when the application initializes with a collection of those interfaces. Any command that needs to update that view would traverse the collection notifying that each view should be updated.
Unlike typical examples the views would be User Controls. You have the flexibility of making any UI element implement that interface so you could use dialogs, full forms, etc in addition to your User Control.
Finally remember the User Control is NOT the view but the implementation of the View. Whatever scheme you adopt you can define what the View as deep as you want and have the User Control implement that interface.
Related
I have view like the image and ViewModel have commands to handle the button (1,2,3,4) clicks. In work area allow the user to give inputs. Depending upon the input users are allowed click the button;
Each Button leads one new Window(View with ViewModel; whose model will be taken from the inputs). The new window have its own logic to showing the data depending upon the model injected to the ViewModel.
As per the MVVM standards, where do I specify the respective View and ViewModels for the each Button? (In View / View model).
How can I specify the same?
Basically ViewModel is a link between View and Model, so each combination of View and Model should have a separate ViewModel (if valid).
In my experience in most cases we are dealing with two kinds of Views:
small views such as icons, advanced buttons and so on (which are more isolated and more likely to have no reference to their parents so they are easy to manage and to be generalized)
large views such as windows, panels (which have a lot of children and are more likely to be changed later)
For small views common ViewModels can be used for multiple Views. but as for large Views (considering possible changes in the future) it's better not to use a single shared ViewModel. however it's helpful to use a base ViewModel class to implement some shared functionality (if any).
So keeping that in mind and focusing to stay standard, I suggest:
ViewModels for secondary windows: (according to the question I think you need 4) Each have their independent functionality (you can derive them all from a BaseWindowVm).best practice here would be not to let them know about their parent (MainWindowVm) and just to set their event handlers when they are instantiated. This way you can avoid code coupling.
MainWindowVm: consists of 4 commands and some other inputs. Each command does these steps:
instantiates a View
instantiates a ViewModel for secondary window based on input
set Vm's event handlers
assign the Vm to DataContext of the View
add the Vm to some list in MainWindowVm (If you want to keep track of these windows)
ShowDialog()
The most important part is that since ViewModels communicate with each other, linking Views with each other only make it more complicated and more difficult to manage. so Views are like islands with bindings to their ViewModels and everything else is up to ViewModels.
With current practices (at least with WPF and Silverlight) we see views bound via command bindings in the view model or we at least see view events handled in view models. This appears to be a violation of SRP because the view model doesn't just model the view state, but responds to the view (user). Others have asked how to build view models without violating SRP or asked whether their implementations do so (this last is the controller in MVC, but roughly analogous).
So are current practices a violation of SRP? Or is "view model" really a collection of things that don't violate SRP? To frame this a bit, it seems we need to know what is the single responsibility or if there are multiple responsibilities in the concept, are the individual responsibilities split out, conforming to SRP. I'm not sure.
Wikipedia's definition of view model says
[T]he ViewModel is a “Model of the View” meaning it is an abstraction of the View that also serves in data binding between the View and the Model
This seems good enough for SRP, but then the entry later says (my emphasis added)
[The ViewModel] acts as a data binder/converter that changes Model information into View information and passes commands from the View into the Model
In a Prism blog post about the view model's role, the author says (again, my emphasis)
What it boils down is that the view model is a composite of the following:
an abstraction of the view
commands
value converters
view state
I'm sure I've missed many definitions out there, but they seem to fall into these categories:
Single "vague" responsibility of modeling view state (so what do we
mean by state)
Multiple responsibilities (view state, user interaction (i.e.
commands))
A composite of single specific responsibilities (abstraction,
state, interaction, conversion), thus having a single
responsibility: "managing all that stuff".
If you're curious, I "care" about this because (2) feels right, but seems counter to the prevailing implementations.
Single Responsibility as Martin defines it:
"THERE SHOULD NEVER BE MORE THAN ONE REASON FOR A CLASS TO CHANGE."
A ViewModel, as far as MVVM is concerned is really just a specialized implementation of a Presentation Model.
So while it could be argued that a Presentation Model should only represent the state of the UI, and that a Presenter/Controller should always broker the commands between the UI and the Presentation Model. If one follows this idea, with SRP dividing on State and Commands, then adding a command should not affect the class that represents state. Therefore MVVM would break SRP.
However...
I think this is grasping at straws. MVVM is a fairly specialized implementation used basically in WPF/Silverlight (and now browser clients).
Patterns are designed to make designs simpler where the alternative would be more cumbersome or less maintainable. Since MVVM is designed to take advantage of the extremely rich data binding capabilities of the presentation technologies, then it is a worthwhile trade off.
No! MVVM does not violate SRP, (the programmer does, lol!)
There is no reason that using the MVVM pattern needs to ignore the SRP. MVVM does not mean that there is only one Model Class, one View-Model Class, and one View Class. Certainly, if you only had one View Class, you could only ever show one simple screen.
Those classes that are in the View tier, should be responsible for one thing; doing, deciding, or containing. A View can consist of several sub-views who's jobs are to do certain pieces of the users interractions. Consider a basic form, it has a display grid, items in the grid can be edited, and there is a "Save" button.
The main View would be a container for two other views; the datagrid (a user control, or something) and a command control. The datagrid then is responsible for choosing the right childview to render data in; in essense it's a container that databinds. The View to edit items is a child view of the datagrid, which is in-turn a child of the main View. Lastly the command control is a set of buttons (in this case a single one) who's single responsibility is to raise signals that commands have been issued by the user.
In this way the Edit View (used by the DataGrid) is agnostic about what uses it, and has one responsibility; Same with the command control. Likewise the DataGrid doesn't care about who uses it, only that it can contain the Edit View (child). Nice SRP there.
ViewModels to match the Views (and children) are also responsible for one thing. The Edit View Model is a container to which the Edit View Binds; it simply contains the data fields that can be displayed/edited. It doesn't care about anything but signalling when one of its properties change. The Command Button View Model is a class that does things. It's commands are bound to the buttons, and it will do work based on what the user clicks on. It will have to have access to other parts of the ViewModel(s) to do it's work.
The main page View Model is there to contain the other child views. It's sole responsibility is as an initializer, making all the required ViewModel instances, and passing constructor parameters to other ViewModel instances (say, the Command Button View Model so it knows where to get data for it's work)
It's natural to cram a whole bunch of functionality into a single ViewModel that a large View would bind to. But it doesn't have to be that way, and SRP can be maintained in MVVM.
The main goal of MVVM is to allow for testable design, the Model layer can be tested independantly, all classes in the Model can easily follow SRP. The ViewModel can be tested without the need of a view; it gets trickier to think SRP in the ViewModel, but it is certainly doable; just remember to break out your classes so they only have one concern. The View is bound to parter ViewModels, with any luck, your testing of the ViewModel makes snapping the View(s) on super easy. Remember you can have each View-let adhere to SRP to be part of a larger conglomerate View (container).
TL;DR?
To answer your question directly, the View is a collection of classes that does not break the SRP. Thus, when the ViewModel is abstracted from the View(s) (View-First), they are also a collection of classes that adhere to good SRP.
I consider many of the current practices around MVVM violate SPR (at least). This is yet another situation where simply adding controllers back to MVVM would solve all the problems cleanly. I call it MVCVM :)
The pattern we are using successfully on all recent projects is to register controllers only, in modules, and initialise them at startup. The controllers are very light/slim and the only thing that needs to hang around for the life of the app listening for, or sending, messages. In their initialise methods they then register anything they need to own (views and viewmodels etc). This lightweight logic-only-in-memory pattern makes for slimmer apps too (e.g. better for WP7).
The problem with just using VMs, as you have found, is that eventually you hit cases where they need to know about views, or commands, or other stuff no self-respecting ViewModel should be involved with!
The basic rules we follow are:
Controllers make decisions based on events
Controllers fetch data and place it in appropriate View Model properties
Controllers set ICommand properties of View Models to intercept events
Controllers make views appear (if not implied elsewhere)
View Models are "dumb". The hold data for binding and nothing else
Views know they display a certain shape of data, but have no idea where it comes from
The last two points are the ones you should never break or separation of concerns goes out the window.
Simply adding controllers back into the MVVM mix seems to solve all the problems we have found. MVVM is a good thing, but why did they not include controllers? (but this is of course just my opinion) :)
What it boils down is that the view model is a composite of the following:
an abstraction of the view
commands
value converters
view state
I don't see why you've separated the first two items. Commands are part of the view.
As for the rest - you're right. In some cases. I've built applications where the tasks of value conversion and maintaining view state were sufficiently complex that it didn't make sense for a single view model class to do it all, and I broke them out into separate classes that interoperate with the VMs.
So?
I'm using MVVM Light toolkit (which I love). I currently have messaging in place for some interaction originating from the ViewModel and intended for consumption by the View. Typically these types of messages indicate the View should do something like hide itself, show a confirmation message that data was saved, etc.
This works. In the constructor for the View, I register with the Messenger:
Messenger.Default.Register<NotificationMessage<PaperNotification>>(this, n => HandlePaperNotification(n));
When I'm using the Messenger to communicate cross-cutting concerns between ViewModels (like identity), I can see that when the ViewModel is cleaned up in the ViewModelLocator, the base class for ViewModels (ViewModelBase) unregisters any subscribed messages. I don't have to do anything, as MVVM Light Toolkit handles that for me. However, when I use them in the Views, I have to expressly unregister them at Closing time, like so:
Messenger.Default.Unregister(this);
I suppose I could implement a base class for Views to inherit from.
However, it strikes me that perhaps this is a code smell to be using the Messenger in the View... it works, but it might not be the best way. I'm wondering if I should instead create a property on the ViewModel and bind whatever part of the View's elements to it. In the example of hiding a form, a property could be a boolean called "Show". As I think about it, I can see that in many cases this will result in having to write a ValueConverter. One way seems less testable. The other way seems to require much more code and perhaps the introduction of excess ValueConverters, which could become a code smell in themselves.
So (after all that build up) my question is this:
Is it preferable to use messages within the View or is it better to add properties (and potentially ValueConverters) to allow the ViewModel to drive it in a more bindable fashion?
In MVVM. ViewModel comunicates with View through DataBinding and Commands. If you need some other functionality, you need to implement it using this means.
Messaging is supposed to be only for ViewModels. Views are supposed to be "stupid" visualisers of your data in ViewModel.
The Messaging logic in MVVM Light is there for communication between ViewModels. I've never run into any communication between View and ViewModel that I couldn't solve with binding and/or commands. Sometimes I need Value Converters and sometimes I need code in the code-behind, but I've never had to make the ViewModel directly push data to the View.
This is an interesting discussion and I found this thread when I was wondering about view model to view communication. Interestingly, MVVMLight's creator seems to find it perfectly acceptable to send messages from a view model to a view. Another example of differing opinions about what is a good MVVM design.
I've got what I think is a fairly simple problem in Silverlight, and I'd like to solve it using MVVM principles, largely as a way of enhancing my own understanding.
Let's say I have a simple LOB app that's designed to let me load up and edit a single Employee (just an example, for the sake of explanation). In my example, Employee is made up of a number of complex objects - an Employee has a ContactInfo, a Skillset, an EmploymentHistory, an AwardsEarned, etc. The idea is that I can have this app load up a single employee and get access to a number of different editor screens. Each component of an Employee has its own editor screen.
Visually, the app just ha a left-hand nav bar and a main view on the right side. The nav bar just lets me type in an employee number and load it into memory as the "active" employee. It has a simple list of links - clicking a link should load the appropriate editor screen on the right side.
There are a couple concepts that I don't think I understand well enough, and I'm having trouble proceeding. I know there's always more than one way to skin a cat, especially when it comes to WPF/Silverlight/XAML/MVVM, but I'm having trouble thinking through all the different concepts and their repurcussions.
View-First or ViewModel First
After much thinking about MVVM, what seems most natural to me is the concept of viewmodel composition that Josh Smith seems to promote in his often-quoted article. It seems like the idea here is that you literally model your UI by composing viewmodels together, and then you let the viewmodels render themselves via typed DataTemplates. This feels like a very good separation of concerns to me, and it also makes viewmodel communication very direct and easy to understand.
Of course, Silverlight doesn't have the DataType property on DataTemplates, to many complaints: one, two. Regardless, what I see promoted much more often than viewmodel composition is a more view-first design, where the viewmodel for the view is typically instantiated in the view's XAML or via a DI container, meaning that you can't hand it any parameters. I'm having a really hard time understanding this: how is a ViewModel supposed to serve a Model to a View if I never get to tell it what data is in the model? Reaching through a view to get to its viewmodel doesn't seem to make sense either. I'm very hazy in this area but it seems the accepted answer "use a mediator/lightweight messaging framework." I'm just going through some tutorials now on the messaging system in MVVMLight and I'll be looking at similar stuff, if for nothing else than simply to understand the concepts, but if anyone can shed some light on this I'd very much appreciate it. Anything involving Unity/Prism or MEF is valid, but I haven't gotten that far in my quest for knowledge yet :-)
Instantiating Views and Selecting Them
Theoretically (I say that because SL doesn't support DataTemplate DataType), the viewmodel composition approach could make this very simple. I could have the right side of the screen be a content control whose Content property is bound to a property called ActiveEditor. A parameterized command for the hyperlinks would set ActiveEditor to a given viewmodel.
With a more view-first approach, how would I proceed with this? The first thing that comes to mind is instantiating all of the controls in the XAML of the main view.
Is manipulating the Content property of a ContentControl a good way to go for this kind of situation, or am I better off doing something like setting visibility on each individual control?
The ViewModel (VM) is written so that it is 'wired up' to the Model but has no knowledge at all of the View - in fact, this is what makes it very good to unit test (see NUnit) as it has no idea, and less does it care, whether it is being used by a UI or a Test Framework.
The VM exposes public CLR properties which implement the ICommand interface, the View, having instantiated a VM using (generally speaking anyway) its default constructor, then binds its Buttons/Hyperlinks etc to these Command properties like. So, for example, you may have a VM that exposes a CloseCommand to exit the app, the View may contain a MenuItem that binds to that command, such as:
<MenuItem Header="E_xit" Command="{Binding Path=CloseCommand}" />
Now, the VM would also expose a public ObservableCollection of objects that you want to display in your UI. Whether you populate this ObservableCollection as part of the VM constructor, or whether you do it via a UI interaction (say another Command assigned to a Button click) is up to you but the end result is that you bind a View control to this exposed ObservableCollection in XAML like:
<y:DataGrid ItemsSource="{Binding Breakdown}"/>
or equivelant for whatever control you are using to display the data (not sure off the top of my head what elements a DataGrid has in Silverlight as opposed to WPF).
Meanwhile...: The Mediator pattern is used for VM's to interact with each other, not for the View to interact with the VM. For example, you might have a custom TreeView that has its own VM on the same View as the main chart screen. In this case you could use a Mediator for the TreeView's VM to communicate with the Charts VM.
As for the last bit of your question, I think set up a basic framework using Josh Smith way in the article you mentioned and use that method to add additional ViewModels to the right hand side of your silverlight app.
Hope that helps a bit at least.
I need some advice on implementing UIs in WPF.
So far, I've been using Code-Behinds, which is extremely easy to get-started, but hell when maintaining/changing/testing.
I've looked at MVP (and its WPF variant - MVVM), but having some trouble getting started.
Assuming I have a UI to build, here's what I think I should do:
1.
Create a "Main UI" Mediator class which specifies ALL high-level operations (LoadSettings(), SetVisibility() ) and events (not triggered by the user, e.g, model data changed) that my UI supports.
2.
Create the "Model" classes to represent the data
3.
Create "ViewModel" classes for my model classes.
4.
For complex behaviours (e.g, a sequence of operations need to be done before the UI can/should update, such as modifying items in a collection), do not rely on ViewModels to update the UI. Instead, do it manually through the Main UI Mediator class.
5.
For simple behaviours (e.g, toggling the visibility/enabled states/etc), use WPF binding to bind the ViewModels' properties directly to the UI.
In this case, the Main UI Mediator class would maintain both the ViewModel and Model objects, and delegate user interactions (to the Model) and UI update requests (to the ViewModel/View) appropriately. The Mediator class also provides a centralised interface which specifies the functionalities of the UI, while acting as a Change Manager (described in GOF's Observer Pattern) to handle complex UI behaviour/reduce redundant UI updates.
Am I on the right track? Should I tweak my approach? Change it completely? At the moment, I lack the experience/knowledge to implement huge/complex UIs, so I don't really know whether I'm on the right track.
Thanks
This is a bit long, sorry about that!
So far, I've been using Code-Behinds, which is extremely easy to get-started, but hell when maintaining/changing/testing.
Yep :) Anytime you have to name a control and write "someControl dot blah" in your code-behind, that's a code smell. It's sometimes unavoidable, but try to limit it as much as possible. Your UI is a projection of the model - ViewModels and ValueConverters are a way to deal with the impedance mismatch between the two domains.
A few problems with your approach:
Create a "Main UI" Mediator class which specifies ALL high-level operations
Instead of doing this, your Window class acts as the "Controller"; the important thing is, use Commanding to define your top-level actions. This way, you can have UserControls decoupled from the Window class, because the UserControl will just call Commands.Open.Execute(null, this), and the Window can handle it, and the UserControl will never explicitly have a dependency on the Window.
Create "ViewModel" classes for my model classes.
In MVVM, the VM part is to help you out - if you can get away with binding directly to the model (i.e. the data doesn't change or you don't mind implementing INotifyPropertyChanged in your models), then doing this (even if you have to use a few IValueConverter classes) is okay. ViewModels are mostly used when the view is so different from the model representation that it'd be ugly to hack up your model, or to "tack on" extra properties that only make sense in this particular view.
while acting as a Change Manager...
Remember that WPF does this for you, via Dependency Properties and INotifyPropertyChanged; don't reinvent the wheel; if you write an OnDataUpdate() function, this is a sign you're not using data binding properly.
e.g, a sequence of operations need to be done before the UI can/should update, such as modifying items in a collection
This is where Commanding is great - your CanExecute function can apply arbitrarily complex logic to decide whether a certain operation can be done, and if you bind it to UI elements like Menus or Buttons, they will automatically disable/enable as needed.
It hasn't been mentioned, but do all of your UI design in XAML.
There is nothing worse than seeing WPF UI's being created via code-behind.