How does an MVC system work? - model-view-controller

I'm trying to learn the MVC pattern, but each place say something different. So now I don't know whats the true MVC.
So I'm guessing its the most pure MVC:
Model is just data and notify data changes.
View reads the messages of the Model to update the view.
Controller reads the user input from View and changes the Model according.
Implementing
Model knows no one.
View knows the Model.
Controller knows both View and Model.
Pseudocode:
/* Model */
class Color{
color = blue;
setColor(color);
notifyUpdate();
}
/* View */
class ColorPicker(model){
model.register(update);
update(){
this.colorToExhibit = model.color;
}
}
/* Controller */
class Colorize(view, model){
view.register(update);
update(color){
model.setColor(color);
}
}
Some questions:
Is that right?
I can't see why the View cannot change the Model directly, but through Controller.
Suppose I have animations to be performed after an action. Who must handle this animation: the Model, the View, or the Controller? Also: the animation logic is part of the Model, View, or Controller? More: Suppose a Poker game. After the user choose an action (say, 'Raise'), the system must play an animation (say, the chips going from player spot to the desk). How can I see this poker example (with animation) as a MVC? Can you explain and give a pseudocode about that?
Thank you.

Model is just data and notify data changes.
View reads the messages of the Model to update the view.
Controller reads the user input from View and changes the Model according.
The Model is more than just data. The model is also the business logic. It contains all of the intelligence of the system, or at least an abstraction of behind-the-scenes intelligence (such as database calls or other service calls). Consider the saying, "Keep your models heavy and your controllers light."
Model knows no one.
View knows the Model.
Controller knows both View and Model.
The Model knows no one, that it correct. The Model should be portable between applications and shouldn't depend on UI concerns in any way. (The View and the Controller are UI concerns in this case.)
The View knows the Model, also correct. The View basically "binds" to the Model. It presents all of the UI elements and places Model data within the UI elements accordingly.
The Controller kind of "knows the View." It knows which View to which it should direct control, but it doesn't know anything about that View. Nor does it know which View from which control previously came. The Controller responds to events. An event comes in from the UI, carrying some kind of state information with it (a ViewModel, perhaps), directs logical control through the Models (where the business logic happens), and responds with a Model (or a ViewModel, if the shape of the data specific to a particular View is different than the Models) and a View.
I can't see why the View cannot change the Model directly, but through Controller.
The View can manipulate the Model within the context of the user interaction, but shouldn't expect those changes to persist in any way. The View should be considered "client-side" and doesn't know anything "server-side." (Even if you're talking about a native application and not a web application.) Persisting any change is considered a UI "action" or "event" and would go to a Controller to make it happen.
Suppose I have animations to be performed after an action. Who must handle this animation: the Model, the View, or the Controller? Also: the animation logic is part of the Model, View, or Controller?
An animation sounds like an entirely UI-based operation. It would be within the View. Is there more happening than just a UI animation? Does the animation change anything in the back-end? For example, if I have a web application and, when a page loads, I want to fade-in some data (an animation)... that's entirely in the View. The data would be delivered to the View like any other data, and the animation takes place entirely within the UI (View). It doesn't do anything from the perspective of the Model or the Controller.
Suppose a Poker game. After the user choose an action (say, 'Raise'), the system must play an animation (say, the chips going from player spot to the desk). How can I see this poker example (with animation) as a MVC? Can you explain and give a pseudocode about that?
The action ("Raise") is a Controller event. The UI would contact the controller to perform the "raise". So the Controller might have a method like this:
View Raise(GameState state)
{
// Interact with the Models to update the known state of the game.
// The Models would perform the actual Poker game logic.
// Respond with a View bound to updated Models.
}
Once the Controller responds to the UI with a new View, that View would contain any animations to display to the user. (After all, you don't want to perform the animation unless the action was successful, right? When the Controller responds to the UI with a new View indicating a successful action, then the animation would play. It may instead respond to the UI with a View indicating an error, in which case that View would show something else.)

I'll go with the simple Bank analogy.
Tellers are Views.
Runners are Controllers.
Bankers are Models.
The Bankers are the smart ones, they know all of the business logic and do all of the complex calculations.
The Runners are used to transport the money (data) from the Bankers to the Tellers.
The Teller presents the money to the Customer.
A simple representation:
Model
public class BankAccount
{
public int ID;
public int Balance;
public BankAccount(int id)
{
ID = id;
Balance = DetermineAmount();
}
public int DetermineAmount()
{
// Gather transaction info, debits, credits and return a
// sum of the amount left in the account depending on the
// id provided.
}
}
Controller
public class BankAccountController
{
public ViewResult Index(int id)
{
BankAccount account = new BankAccount(id);
return View(account);
}
}
View
<ul id="account-info">
<li>Account ID: `#Model.ID`</li>
<li>Balance: `#Model.Balance`</li>
</ul>

If you are interested in the historical true MVC, then start with Trygve Reenskaug. He created (observed?, catalogued??) it in the late 1970s. For a start, read "Models-Views-Controllers" from 1979. It defines the terminology. Take careful note of it's title - all three roles are pluralized. This is the first thing that most people seem to get wrong.
The best description I've found of the original use of MVC is actually in a presentation dated 2004 entitled "Inside Smalltalk MVC". I would guess that the canonical papers that describe the Smalltalk 80 final version of MVC are Krasner & Pope's "A Cookbook for Using the Model-View-Controller User Interface Paradigm in the Smalltalk-80" and
Steve Burbeck's "Applications Programming in Smalltalk-80: How to use Model-View-Controller (MVC)". Both papers are well worth the read.
If you have some time to kill and don't mind listening to Robert Martin, he did a good keynote at Ruby Midwest 2011 that touched on MVC. It is a little over an hour, but quite entertaining and enlightening. I tend to follow with his opinion that most implementations get MVC wrong. I spent a little time looking around and finally found a diagram that I can link to which describes MVC. The one that I like came from Pope and Krasner.
(source: as3dp.com)
From my point of view, the following are the key points:
a model instance is responsible for notifying the interested objects of changes. Note that these can be any object instances. The diagram shows both views and controllers receiving updates here.
views are responsible for querying the current state and displaying the results. They usually perform filtering or data transformation as well.
controllers are responsible for accepting user input and forwarding view messages along to the view.
View messages are a common theme in MVC. It is important that these are independent of the UI world - these are not mouse clicks and what not but a view-specific language of events. This brings us to the next point.
The view does not depend on the controller in any way. Controller's are responsible for arranging and creating views and providing the interface between the rest of the world and the view.
In a perfect world, the view is responsible for making the model representation visible. This is how it worked when MVC was applied to desktop applications.
The reality is that MVC has been twisted and rewritten for the web world. It's not really MVC anymore or maybe MVC was simply redefined. This is why you are seeing so many different opinions and representations of MVC out there. If you are looking into writing desktop style applications, then look at the stuff by Krasner & Pope. If you are looking into how MVC is applied to the web, then I recommend Uncle Bob's keynote for an alternative that it better suited for web applications - what he called the Interactor, Entity, Boundary Architecture for lack of a better name. Dig around for stuff associated with his talks about "the Lost Years of Architecture".

Related

The point of MVC

First, just to say, this is not a troll posting, I'm genuinely completely and utterly at a loss as to why MVC is where the industry is going/gone. If MVC wasn't on every job advert I've seen for the last 4 years I'd not bother but as it is, it looks like I'm going to have to use it.
What is the point of MVC? Yes you get the pretty URL e.g. www.amazon.co.uk/books/horror/ but I can't see any other advantage. I was reading the ScorrGu blog where other's where asking the same
https://weblogs.asp.net/scottgu/introducing-razor
This is the copy+paste reasoning from the ScottGu blog concerning Razor
Some people like a controls model with code-behind for code/content separation, and others like a MVC model where there is separation between the controller and view. In an MVC model your application/business logic is contained within the controller and model layers - and does not live in your view. Your view file instead just focuses on generating HTML rendering.
In the controls model there is "separation between the controller and view" in that there is the code behind(controller) and the content (view).
Application/business logic is sprinkled throughout the view (depending on what's on that view obviously) from the examples and tutorials I've been looking at logic like "Does this user have permissions to see this link?"
"Your view file instead just focuses on generating HTML rendering."
It isn't
#foreach (var item in Model.Items) {
var style = item.IsArchived ? "archived-validation-item" : "";
#Html.DisplayFor(modelItem => item.Code)
their is logic here about how many times a loop is performed and retrieving the right item code.
In the controls model, a page has two files one containing the logic and one the layout.
In MVC, a page has three files, the controller and model containing logic and the view which contains layout and logic.
The reason classic ASP was left behind was so that the code and the markup were seperated, the front end developer could get on with making it look pretty while the back end developer worked on getting the right things onto that page. Their roles and responsibilities were clear and seperate. With MVC, both developers must know about the model (and viewmodel too if that's used) as well as the view and controller (as well as the routes, etc.) so that everything is linked up together. From the examples I've seen for example, the decision on when a link should be shown could be made by the view, the controller or the model. This adds layers of complexity and confusion.
What am I missing apart from some well paid jobs? ;)
Thanks
You are correct in that the old *.aspx + Code-Behind Class was an example of separating concerns and logic - however it was not true MVC separation because the code-behind class was still concerned with aspects of presentation, and similarly it's possible for the .aspx file to have business logic in it, as an example, this:
ASPX:
<!-- this data-source is arguably a piece of business logic in the view layer: -->
<asp:SqlDataSource runat="server" id="ds1" selectCommand="SELECT * FROM someTable" />
<asp:GridView runat="server" id="gv1" dataSource="ds1" />
<asp:Literal runat="server" id="lit1" />
Code-behind:
void Page_Load(Object sender, EventArgs e) {
this.lit1.Text = "<div class=\"something\"><p>some HTML goes here</p></div>"; // this is presentation logic in the code-behind
}
In MVC separation, you remove the "tight-coupling" between the two, as well as enforcing the separation of concerns because you cannot (feasibly) generate HTML from within a Controller Action, and similarly you should not be able to execute business rules in a View because for example, all database connections are closed at that point in the request lifecycle.
One core tenet of MVC is the "view-model" which is a class that encapsulates the entirety of data represented (and manipulated) by the view (alternatively, the tuple of the ViewModel with the ViewData). As this class is well-defined it means the page is almost self-describing, there is no need to manually examine the view to see what's going on and what data is being displayed.
In summary, it's this:
WebForms:
[query database] -> [populate view directly]
MVC:
[query database] -> [populate viewmodel] -> [view renders viewmodel]
One advantage to this approach is that you can turn a website into a web-service by only changing one line of code:
From this:
public ActionResult Index() {
IndexViewModel viewModel = new IndexViewModel();
// (populate ViewModel here)
return this.View( viewModel );
}
To this:
public ActionResult Index() {
IndexViewModel viewModel = new IndexViewModel();
// (populate ViewModel here)
return this.Json( viewModel );
}
You can't do that with WebForms :)
You can write good code, or bad code independent of the frameworks involved. IMHO MVC is a pattern that better lends itself to test-ability due the view separation. Putting logic in the correct places is still the responsibility of all the team members. MVC won't guarantee a great architecture.
And honestly if your application is correctly designed and your webforms are all calling logic in services, maybe MVC isn't that big of a deal for you personally. Also, I agree with what Vincent mentioned about the shift toward SPA's. Personally I've seen several applications written as spas that aren't interactive that would have been just fine as MVC apps. As always, "It depends" is still the best software dogma.

Is Ext JS's MVC an anti-pattern?

I work in a team of 25 developers. We use ExtJS MVC pattern of Sencha. But we believe that their definition of MVC is misleading. Maybe we might call their MVC an anti-pattern too.
AMAIK, in MVC controller only knows the name or the path of the view, and has no knowledge on the view's internal structure. For example, it's not the responsibility of the controller, whether view renders the list of customers a simple drop down, or an auto-complete.
However, in Ext JS's MVC, controller should know the rendering of view's elements, because controller hooks into those elements, and listens to their events. This means that if an element of the view change (for example a button become a link), then the relevant selector in the controller should change too. In other words, controller is tightly-coupled to the internal structure of the view.
Is this reason acceptable to denounce Ext JS's MVC as anti-pattern? Are we right that controllers are coupled to views?
UPDATE (March 2015): Ext 5.0 introduced ViewControllers that should address most of the concerns discussed in this thread. Advantages:
Better/enforced scope around component references inside the ViewController
Easier to encapsulate view-specific logic separately from application flow-control logic
ViewController lifecycle managed by the framework along with the view it's associated with
Ext 5 still offers the existing Ext.app.Controller class, to keep things backwards-compatible, and to give more flexibility for how to structure your application.
Original answer:
in Ext JS's MVC, controller should know the rendering of view's elements, because controller hooks into those elements, and listens to their events. This means that if an element of the view change (for example a button become a link), then the relevant selector in the controller should change too. In other words, controller is tightly-coupled to the internal structure of the view.
I actually agree that in most cases this is not the best choice for the exact reasons you cite, and it's unfortunate that most of the examples that ship with Ext and Touch demonstrate refs and control functions that are often defined using selectors that violate view encapsulation. However, this is not a requirement of MVC -- it's just how the examples have been implemented, and it's easy to avoid.
BTW, I think it definitely can make sense to differentiate controller styles between true application controllers (control app flow and shared business logic, should be totally uncoupled from views -- these are what you're referring to), and view controllers (control/event logic specific to a view, tightly-coupled by design). Example of the latter would be logic to coordinate between widgets within a view, totally internally to that view. This is a common use case, and coupling a view-controller to its view is not an issue -- it's simply a code management strategy to keep the view class as dumb as possible.
The problem is that in most documented examples, every controller simply references whatever it wants to, and that's not a great pattern. However, this is NOT a requirement of Ext's MVC implementation -- it is simply a (lazy?) convention used in their examples. It's quite simple (and I would argue advisable) to instead have your view classes define their own custom getters and events for anything that should be exposed to application controllers. The refs config is just a shorthand -- you can always call something like myView.getSomeReference() yourself, and allow the view to dictate what gets returned. Instead of this.control('some > view > widget') just define a custom event on the view and do this.control('myevent') when that widget does something the controller needs to know about. Easy as that.
The drawback is that this approach requires a little more code, and for simple cases (like examples) it can be overkill. But I agree that for real applications, or any shared development, it's a much better approach.
So yes, binding app-level controllers to internal view controls is, in itself, an anti-pattern. But Ext's MVC does not require it, and it's very simple to avoid doing it yourself.
I use ExtJS 4's MVC everyday. Rather than spaghetti code, I have an elegant MVC app that has tightly defined separation of concens and is ridiculously simple to maintain and extend. Maybe your implementation needs to be tweaked a bit to take full advantage of what the MVC approach offers.
Of course the controllers are bound to the views in some way. You need to target exactly which elements in your views you want to listen to.
eg: listen to that button clicks or to that form element change or to that custom component/event.
The goal of MVC is components decoupling and reusability and the Sencha MVC is awesome for that. As #bmoeskau says, you have to be careful in separation of the view controllers (builtin for the view/widgets itself) and the application controllers (top level views manipulations) to take full advantage of the MVC pattern. And this is something not obvious when your read http://docs.sencha.com/ext-js/4-1/#!/guide/application_architecture. Refactor your MVC approach, create different controllers, create custom component, and embrace the full ExtJS MVC architecture to take advantage of it.
There's still a slight problem in Sencha approach IMHO, the MVC refs system doesnt really work when you have multiple instances of the same views in an Application. eg: if you have a TabPanel with multiple instances of the same Component, the refs system is broken as it will always target the first element found in the DOM... There are workarounds and a project trying to fix that but i hope this will be adressed soon.
I'm currently undergoing the Fast Track to ExtJS 4 from Sencha Training. I have a strong background in ExtJS (since ExtJS 2.0) and was very curious to see how the MVC was implemented in ExtJS 4.
Now, previously, the way I would simulate kind of a Controller, would be to delegate that responsibility to the Main Container. Imagine the following example in ExtJS 3:
Ext.ns('Test');
Test.MainPanel = Ext.extend(Ext.Container, {
initComponent : function() {
this.panel1 = new Test.Panel1({
listeners: {
firstButtonPressed: function(){
this.panel2.addSomething();
},
scope: this
}
});
this.panel2 = new Test.Panel2();
this.items = [this.panel1,this.panel2];
Test.MainPanel.superclass.initComponent.call(this);
}
});
Test.Panel1 = Ext.extend(Ext.Panel, {
initComponent : function() {
this.addEvents('firstButtonPressed');
this.tbar = new Ext.Toolbar({
items: [{
text: 'First Button',
handler: function(){
this.fireEvent('firstButtonPressed');
}
}]
});
Text.Panel1.superclass.initComponent.call(this);
}
});
Test.Panel2 = Ext.extend(Ext.Panel, {
initComponent : function() {
this.items = [new Ext.form.Label('test Label')]
Test.Panel2.superclass.initComponent.call(this);
},
addSomething: function(){
alert('add something reached')
}
});
As you can see, my MainPanel is (besides the fact that is holding both panels) also delegating events and thus creating a communication between the two components, so simulating sort of Controller.
In ExtJS 4 there is MVC directly implemented in it. What really striked me was that the way the Controller actually fetches the components is through QuerySelector which in my opinion is very prone to error. Let's see:
Ext.define('MyApp.controller.Earmarks', {
extend:'Ext.app.Controller',
views:['earmark.Chart'],
init:function () {
this.control({
'earmarkchart > toolbar > button':{
click:this.onChartSelect
},
'earmarkchart tool[type=gear]':{
click:this.selectChart
}
});
}
});
So as we can see here, the way the Controller is aware of the earmarkchart button and tool is through selectors. Let's imagine now that I am changing the layout in my earmarkchart and I actually move the button outside of the toolbar. All of a sudden my application is broken, because I always need to be aware that changing the layout might have impact on the Controller associated with it.
One might say that I can then use itemId instead, but again I need to be aware if I delete a component I will need to scatter to find if there is any hidden reference in my Controllers for that itemId, and also the fact that I cannot have the same itemId per parent Component, so if I have an itemId called 'testId' in a Panel1 and the same in a Grid1 then I would still need to select if I want the itemId from Panel1 or from the Grid1.
I understand that the Query is very powerful because it gives you a lot of flexibility, but that flexibility comes at a very high price in my opinion, and if I have a team of 5 people developing User Interfaces and I need to explain this concepts I will put my hands on the fire that they will make tons of mistakes because of the points I referenced before.
What's your overall opinion on this? Would it be easier to just somehow communicate with events? Meaning if my Controller is actually aware of what views he's expecting events, then one could just fire an event dosomethingController and the associated Controller would get it, instead of all this Query problem.
I think if you use the Sencha Architect to produce the Views then Inherit from that View to create Your own View.
Now this View Can be responsible to hook up to any events and raise meaningful events.
This is just a thought...
//Designer Generated
Ext.define('MyApp.view.MainView', {
extend: 'Ext.grid.GridPanel',
alias: 'widget.mainview',
initComponent: function() {
}
});
//Your View Decorator
Ext.define('MyApp.view.MainView', {
extend: 'MyApp.view.MainViewEx',
alias: 'widget.mainviewex',
initComponent: function() {
this.mon(this, 'rowselect', function(){
this.fireEvent('userselected', arguments);
}, this);
}
});
I think there is a pretty bad problem here - its very difficult to shard isolated units within a page.
The approach I'm experimenting with (which makes it somewhat easier to write tests aswell) is to have a vanilla js context object for each page which contains the logic (and has the benefit of being very easy to break up and delegate to different objects). The controllers then basically call methods on the context objects when they receive events and have methods on them for retrieving bits of the view or making changes to the view.
I'm from a wpf background and like to think of the controllers as code-behind files. The dialog between presenter/context and view is a lot chattier than wpf (since you dont have binding + data templating) but its not too bad.
Theres also a further problem I haven't had to solve yet - the use of singletons for controllers causes problems for reuse of UI elements on the same page. That seems like a serious design flaw. The common solution I've seen is (yet again) to pile everything into one file and in this case to ditch the controller altogether. Clearly that's not a good approach as soon as things start to get complicated.
It seems like getting all the state out of the controller should help though and you'd then have a second level of context objects - the top level one would basically just assign a unique id to each view and have a map of context=>view and provide dispatch to the individual context methods - it'd basically be a facade. Then the state for each view would be dealt with in the objects dispatched to. A good example of why statics are evil!

Prism navigation- finding out origin of navigation request?

I have a viewmodel tied to a view used in a region. I'm trying to find a way that when that view is navigated to from a particular view (say view A), it does some work internally, like initializing some lists, setting some stuff, whatever. But if it has been navigated to from view B, it needs to NOT reinitialize everything, and just display the data it already has.
I could pass a parameter I suppose, saying whether this is a new operation or if we are going back to work on the old one, but I thought it would be nicer to be able to state that if we came from this view, we do one thing, and if we came from that one we do another.
If that makes sense :)
You can implement the INavigationAware interface which contains 3 methods. One of these methods is the OnNavigatedTo method. There you can access the journal and check the current entry. From there you should be able to determine if it came from View A or View B.
public void OnNavigatedTo(NavigationContext navigationContext)
{
var journal = navigationContext.NavigationService.Journal;
//use journal.CurrentEntry
}

MVVM Light: Where do I instantiate a class from a model and how do I update/access it?

Right now, I'm working on my first WP7 app and have run into some questions, which I haven't been able to answer despite reading what I could find online. Please consider an app that has a main page, a parameters page and a results page. In the parameters page, the user can enter or update numbers in various textboxes. Hitting the back button takes the user back to the main page, where there is a button called "Calculate". Hitting that button should take the data, perform a calculation with it and take the user to the results page presenting a grid with the results.
In a file called Calculator.cs I have a class called Calculator inside a folder called Models. I also have my MainViewModel.cs, ParametersViewModel.cs, and ResultsViewModel.cs files inside the ViewModels folder and the corresponding MainPage.xaml, along with Parameters.xaml and Results.xaml inside a folder called Views. I'm assuming that all the data will be manipulated within the instance of the Calculator class and then a results set will be returned and directed to Results.xaml. I'm just at a loss as to where to instantiate the Calculator class, pass it data, then retrieve the results. I'm also somewhat puzzled how I will trigger the automatic navigation to the Results page when the calculation is done.
Any help with this would be greatly appreciated.
UPDATE: Passing a complex object to a page while navigating in a WP7 Silverlight application has some more info on the same subject. I can go into App.xaml.cs and add something like this:
public class Foobar
{
public string barfoo = "hah!";
}
public static Foobar myfoob = new Foobar();
Then access it from a ViewModel page, e.g. AboutViewModel.cs, like this:
public AboutViewModel()
{
string goo = App.myfoob.barfoo;
}
But at this point I'm still uncertain what unforseen effects that might have. I'm going to tackle serialization/tombstoning at this point to see what happens with either this approach or by using the same DataContext across pages. Otherwise, one of the posters in the link above mentioned serializing the params and passing them between pages. My concern there would be whether or not there is a character limit as with HTTP GET. Seems there is: URI Limits in Silverlight
There are of course lots of possible designs - and lots of them are correct in different ways!
Here's one I might use:
The Calculate button press should trigger the Navigate to the Results page
On navigate to, the Results page should show some animation (maybe just a progress bar)
On navigate to, the Results page should create a new ResultsViewModel, passing in the MainViewModel as parameters
the constructor (or some init method) of the ResultsViewModel should spark up a thread to do the calculation
when this calculation is complete, then the relevant properties of the ResultsViewModel will get set
at which point the databinding on the Results page will clear the animation and show the results
Other solutions are definitely available - will be interested to read what other people suggest and prefer.
As an aside, one thing to watch out for on your Results page is tombstoning - could be an interesting challenge!

MVC using ASP.NET webforms

I've inherited an ASP.NET application built on top of webforms, and the application suffers from having all of its business logic embedded in the codebehind. As a result, this application can't be unit tested.
I want to break out the functionality of every form into an MVC style, but I've found that ASP.NET resists every effort on my part to refactor it. In general, I like to seperate my MVC classes as follows:
public class LoginModel
{
public string Username, Password;
public bool IsAuthenticated;
}
public interface ILoginView
{
event Action UserLoggedIn;
void SetMode(bool isAuthenticated);
}
public class LoginController
{
ILoginView View;
LoginModel Model;
public LoginController(ILoginView view, LoginModel model)
{
this.View = view;
this.Model = model;
// hook onto view events
}
}
Once I've got my classes set up and nicely unit tested, I can implement the ILoginView interface on my usercontrol or page:
public class LoginView : UserControl, ILoginView
{
public LoginView() : base()
{
new LoginController(this); // registers view with the controller
}
}
If this were a winform application, it would work beautifullly. But the ASP.NET lifecycle causes this style to break down.
ASP.NET creates and destroys the view on every page load. Since my Controller is held by the View, the Model is held by the Controller, every postback causes my page to lose its state.
I can remedy the problem above by holding my Controller in the users session, but that introduces a world of problems. In particular, putting Controllers in the session causes memory issues because Models and Controllersaren't reclaimed by garbage collection until the session expires. Navigating from page to page creates dozens of controllers, but the controllers don't dispose of themselves when the user navigates away from a page.
Since a view is destroyed/recreated on every postback, then I have to re-register the view with the controller on every postback. This is harder to do than it sounds, because the state of the Model needs to be copied back to the View on every postback, but simultaneously we don't want to overwrite a users changes to the View which were made in the previous postback. You have no idea what kind of additional nightmare this turns into when dealing with dynamically created or AJAXed controls using this MVC style.
I know I'm overthinking this and there is an easier way to get the results I want, but how do I implement an MVC style properly using webforms?
Is it not easier to re-write this using asp.net mvc?
As it seems you'll have to re-write. Either completely to MVC (and stop using DNN), or to a better WebForms implementation (meaning, separate logic from display and take into account the page life-cycle issues). There's a third option - combining MVC and ASP.NET WebForms, but you should look at it carefully taking into account all the variables in your platform.
Well the nature of the POSTback is dictating your state changes, so you should react on that. Any framework you use would work pretty much the same way, it'll rebuild/bind the state with each request. You should look into savind the state (read data) into your user's session.
ViewState should be used to store all the loaded data in WebForms, so the Controller would be instancied only when the page is created, eliminating the need to store any object in the user session.

Resources