I am a beginner with Blender and I searched about working with actions but have not found the solution to my problem.
I work with a file containing many actions where each action represents a specific body movement, using an armature. Here is my workflow:
Create new action.
Create fake user.
Move the armature and save a few keyframes "LocRot".
Push down action.
Repeat with next action.
Then I have a script that renders all the actions in one go.
The problem: sometimes after adding a new action or a new object, other actions are affected and move to another location. What could cause that?
Also: Is it possible to make some objects (other than the main object's armature) appear on only specific actions? And either make them disappear or go out of the camera view for the other actions.
For example, there is a ViewModel that has
a collection of objects requested from a Service
an Add command to create and add new objects
a Delete command to delete selected objects
The User creates, adds and deletes objects by using corresponding commands.
How should ViewModel react on User's actions?
Should ViewModel immediately make a corresponding Service method call when the User invokes a command?
OR
Should ViewModel add / remove items from local collection, mark items as added / deleted and wait to make a Service method call only when the User invokes some Save / Apply changes button?
The second variant seems more complicated.
Does it give any performance / user experience benefit or enable any features that are not possible with the first variant?
In other words, is it better to keep ViewModel state as close to the Service / Database state as possible?
I have a core data 'ShoppingList' which contains 'Item' objects. I store a display order as an attribute of each item.
I would like to update the display order of all other items in the shopping list whenever an item is deleted. The code to do this is working fine when I use it in my view controller (from where the item is deleted), but since it is really related to the business objects and not the view, it would be better placed in either ShoppingList or Item.
Ideally, I would like it incorporated into the deletion of the item. So far I have tried the following:
1) Customize the standard Core Data generated ShoppingList.RemoveItemsObject (making sure to observe KVO before.after). What's strange about this way is that the item passed is stripped of its relationships to other core data entities before it gets to my code, which I need to process display orders correctly.
2) Customize Item.didTurnIntoFault. Same applies - but even attributes of the item are gone by this stage.
One answer would be to simply define a new method on ShoppingList that does my processing and then calls the original removeItemsObject. But I would prefer to know that whenever an item is removed, from anywhere, this is taken care of. This works nicely when I customize awakeFromInsert, for example - I know that whenever an item is created certain things are setup for me. But I'm surprised there's no equivalent for deletion.
Did you try to implement prepareForDeletion? Sounds like it's exactly what you're looking for.
The doc says:
You can implement this method to perform any operations required before the object is deleted, such as custom propagation before relationships are torn down, or reconfiguration of objects using key-value observing.
I have encountered a problem with the MVC pattern in ExtJS 4. At least, I think I have. Having approached multiple people with this question, and having posted numerous times in the Sencha forums, I am now turning to a broader audience in hopes of getting either a light bulb or a confirmation.
Problem
Your application has the ability to open many different views, some of which themselves are mini-applications. Additionally, a user may wish to have multiple concurrent copies of a view open.
This application is a single-page client-side Javascript application.
The ExtJS 4 MVC model expects you to define all of your controllers in your Application class. These controllers are then initialized when the Application loads. Controllers keep track of views, models and stores.
When you initialize controller A multiple times, say to create more than one copy of a view, you end up with two views that reference the same data stores, and functionally send duplicate events to the Application event bus.
I have refactored my application by adding new prototype methods to Component and Controller to allow for both a) sub controllers (some of my controllers were getting pretty huge) and b) defining stores specifically for the view they work with. The models can still be defined on the controller, just for ease of use by handlers if you need to do something like grab a record from the server.
Question
My understanding of MVC would lead me to believe that models more directly relate to the View than then Controller. I asssssume that ExtJS 4 decides to attach stores (which I think can be seen as wrappers to a more classic model) to Controllers for purposes of encouraging re-use of loaded data, and to optimize away from having many copies of the same class instantiated. It seems to me, however, that one cannot do this if one intends to have many instances of a view available to the user. To my thinking, having many instances is an important option in an OO framework, hence why I have bucked the trend and implemented prototypes on some of the Ext base classes. (Thank you Ext.implement!).
Is there any way to have multiple concurrent instances of a view with different data loaded into them using the out of the box MVC classes and making uses of the provided setters, getters, etc?
I was faced with a similar problem:
Consider a tabpanel for a CRM type application which opens new instances of a view for each client. And say that tab view contains 3 or 4 row-editing gridpanels for interacting with different collections of data relating to that client.
The solution I came up with was based on this from the Sencha forums. In a nut shell, almost all events that are dispatched from a view contain a reference to the view itself. The handlers in my controller's control function all use these to get a reference to the correct view instance.
For dealing with the multiple instances of the same store needed for this, I took this to heart from that post:
For the Store instance on the view or a global one... depends on the
needs. If you are going to use globally then make it global. If you
only are going to need it on the view then put it on the view. MVC is
not a law, you can change it to fit your needs. Technically the
Controller part of MVC is suppose to be the middle man between the
View and Model parts but sometimes that's just not needed. I create
the Store in the view 95% of the time. I'll give you an example...
If you have a Store for products, you probably only need to reference
that Store in your Grid. That usually isn't needed for other parts of
the application. However, if you have a Store to load countries, I
often need it globally so I only have to load it once and can then
set/use that Store in several views.
So I just created the needed store's that relate to a view instance specifically, inside the view's initComponent method. The application did have a few global stores that I created as store classes following the MVC recommendations. It worked out nicely to encapsulate the view instance stores inside the view. Then I only needed one instance of the controller.
To answer your question specifically, currently, there is no ExtJS official recommendation or config for dealing with multiple instances of the same view that use the same store constructor. I have spent some time looking for something like that and the best I have found was this recommendation from one of their forum moderators.
I don't think you ever need more than 1 instance of a controller, regardless of how many views/models you have. See functional example here:
http://whatisextjs.com/extjs-4-extension/fieldset-w-dynamic-controls-7
This can be done, reasonably easily. You need to follow a few rules:
load your controllers at app startup. Don't unload them. Don't worry about the memory or time, it's pretty small even for hundreds of controllers, as long as you minimize and concatenate your js.
Never use the refs or views properties of a controller. You are going to use one instance of a controller, but multiple instances of views, so you never want a reference to a view.
only use event listeners in controllers. You are only going to listen to events on your views. You can always get a (temporary) reference to a view in the event handler via the "cmp" parameter in the handler.
To "launch" a view, create it and add it to another view. To destroy it, destroy it. You don't use a controller to launch a view. You can use the afterrender and beforedestroy events in the controller to add logic.
In ExtJS' MVC the controller is a Singleton for you view. I like how DeftJS thinks about MVC. Each instance of a view has an own instance of a controller. In this way you can put all "controlling rules" in a controller for a particular part of your view, and this will be instantiated only when the view opens.
I did not have any experience how I could use multiple Defts JS apps in the same project.
Of course. What led you to believe otherwise?
Here is an example of creating a custom View which extends from a Window component. You can run this method many times from the same controller and each time you will get a new instance of a View.
"this" refers to a controller that code runs in:
this.getRequestModel().load(requestID,{ //load from server (async)
success: function(record, operation) {
var view = Ext.widget('requestEdit',{
title: 'MyRequest '+requestID
});
var form = view.down('form');
form.loadRecord(record);
}
});
How do you create your views? I see no reason why you cannot pass a different store or config data to every object. Some code samples would help for what exactly you are doing. For example, we have a similar sounding application, and everything is done with extensions. So, if we need a grid, we run
Ext.define('MyApp.grids.something',{
extends:'Ext.grid.panel'
//...
These classes are predefined. Then, when a controller or view is loading this grid, they are using
var grid=Ext.create('MyApp.grids.something',{id:'unique',store:mystore});
As you can see, we can pass in different config options to the same grid each time it is created. We can treat this exactly as you would treat
Ext.create('Ext.grid.Panel');
Except of course that we make some options predefined, and some non-override-able, and so on.
Hope this helped.
Check out this post. The idea there is to take some configuration (like store and itemId) from view config and put it into the viewport config:
// .../app/view/Viewport.js
Ext.define('MyApp.view.Viewport', {
// ...
items: [
// ...
{ xtype: 'testview', store: 'Store1', itemId: 'instance1' },
{ xtype: 'testview', store: 'Store2', itemId: 'instance2' }
]
});
The problem with store will be solved, obviously. Different itemIds will enable you to handle events properly.
I'm using a sheet to ask the user some information for an operation that the application performs. The sheet has some 10 to 15 options (split into 4 tabs, so its a relatively clean UI) that the rest of the program needs to know before proceeding.
Right now, I have a separate window controller class for the sheet called SheetController. SheetController has a delegate property and the main controller, AppController, is set as the delegate.
When the user clicks OK in the sheet the delegate is notified and a method called didClickDone:(id)sender withParameters:(id)parameters is executed. parameter is a object that contains various properties from the sheet. My question is, is this a good approach to handling Sheets and returning data from them?
Secondly, one thing that bothers me is that the parameter is just a dead object - it only has accessor methods. It doesn't do any manipulation on them as its entire purpose is to 'carry' the data to the main controller, which in turn passes the information to the Model of the program. And, unless I'm missing something, shouldn't I just declare parameter to be a normal C struct? Or is there an advantage of using an object for this sort of purpose?
EDIT: Would a passing an NSDictionary be a good compromise? I could save the returned information with their keys in the dictionary and simply pass it.
Maybe I would not notify directly the controller from the push button event handler but
call another routine that encapsualates the logic on how the data have to be commited to the model.
Don't worry about the empty object. Is a common pattern to use it and is called DataTransferObject