If you have read my previous posts on the topic, you know I’m a big fan of custom markup extensions… However, they have a limitation that can be quite annoying : they are only evaluated once. Yet it would be useful to be able to evaluate them again to update the target property, like a binding… It could be useful in various cases, for instance : if the value of the markup extension can change in response to an event

Today I’d like to present an idea that occurred to me recently. Nothing about WPF this time, this is all about C# class design ! The problem It’s very common in C# programs to have an object that owns a collection of child items with a reference to their parent. For instance, this is the case for Windows Forms controls, which have a collection of child controls (Controls), and a reference to their parent control (Parent).

Here’s a piece of code I wrote a while ago, I just thought it could be useful for WinForms developers… In WPF, there is a very handy method to move a window with no borders : Window.DragMove. It can be used like that : private void Window_MouseDown(object sender, MouseButtonEventArgs e) { this.DragMove(); } When you call this method, the window is moved with the mouse until the button is released. It could hardly be simpler ;)

As you may have noticed, it is not possible to modify the contents of an ObservableCollection on a separate thread if a view is bound to this collection : the CollectionView raises a NotSupportedException : This type of CollectionView does not support changes to its SourceCollection from a thread different from the Dispatcher thread To illustrate this, let’s take a simple example : a ListBox bound to a collection of strings in the ViewModel :

It’s quite simple, in WPF, to present data in a grid, thanks to the GridView class. If you want to sort it, however, it gets a little harder… With the DataGridView in Windows Forms, it was “automagic” : when the user clicked a column header, the grid was automatically sorted. To achieve the same behavior in WPF, you need to get your hands dirty… The method recommended by Microsoft is described in this article ; it is based on the Click event of the GridViewColumnHeader class.

If you develop WPF applications according to the Model-View-ViewModel pattern, you may have faced this issue : in XAML, how to bind a key or mouse gesture to a ViewModel command ? The obvious and intuitive approach would be this one : <UserControl.InputBindings> <KeyBinding Modifiers="Control" Key="E" Command="{Binding EditCommand}"/> </UserControl.InputBindings> Unfortunately, this code doesn’t work, for two reasons : The Command property is not a dependency property, so you cannot assign it through binding InputBindings are not part of the logical or visual tree of the control, so they don’t inherit the DataContext A solution would be to create the InputBindings in the code-behind, but in the MVVM pattern we usually prefer to avoid this… I spent a long time looking for alternative solutions to do this in XAML, but most of them are quite complex and unintuitive.

You probably noticed that, in a C# project tree, some items are placed “under” a parent item : it is the case, for instance, for files generated by a designer or wizard : The following trick shows how to apply the same behavior to your own files. Let’s assume that you want to customize the classes generated by the EDM designer. You can’t modify the Model1.designer.cs file, because you changes would be overwritten by the designer.

Soon after the release of WPF, people have been talking more and more about “Model-View-ViewModel” (MVVM). This expression refers to a design pattern, drawing its inspiration from the Model-View-Controller (MVC) and Presentation Model (PM) patterns, and created specifically to take advantage of WPF features. This patterns enables an excellent decoupling between data, behavior and presentation, which makes the code easier to understand and maintain, and improves the collaboration between developers and designers.

Today, I stumbled upon a very handy class : SyndicationFeed. This class, introduced in .NET 3.5, allows to manipulate syndication feeds (like RSS 2.0 or Atom 1.0) with very little code. It can be used to create and publish our own feeds, or to read existing ones. For instance, here’s how to retrieve the news feed from Google News and display its title, its hyperlink, and the titles of it’s items :

Oops… 2 months already since my previous (and first) post… I really have to get on a more regular schedule ;) If you’ve ever tried to use the Clipboard.GetImage method in WPF, you probably had an unpleasant surprise… In fact, this method returns an InteropBitmap which, in some cases (most cases actually), can’t be displayed in an Image control : no exception is thrown, the image size is correct, but the image either appears empty or unrecognizable.