By:
On:
Tagged: , , , ,

There’s an issue I’ve encountered in every single of my UWP applications, so it’s very likely you know about it. In Universal Apps, you can subscribe to the SystemNavigationManager.GetForCurrentView().BackRequested event. It’s usually done firsthand to mimic Windows Phone’s behavior where pressing the back button loads the previous page in the back stack. So far, so good. But then comes the moment when you need to implement a back button behavior that is specific to a page. For instance, you have a popup and you want to dismiss it when the user presses the back button. So you subscribe once more to the SystemNavigationManager.GetForCurrentView().BackRequested event, implementRead More →

By:
On:
Tagged: , , , ,

There’s a dilemma you probably faced if you’ve written a mobile application displaying a large number of pictures. You could display gorgeous hi-resolution pictures, but if you do so, you’ll provide a poor experience to users who use your app on the go. Particularly, users with a poor mobile connection who will have to wait minutes before the pictures are displayed. You may also discontent users with limited data plans, who don’t want to eat gigabytes of data just to display your pictures. On the other hand, if you settle for low-resolution pictures, you’ll provide a sub-optimal experience you are at home, with an unlimited andRead More →

By:
On:
Tagged: , , ,

I came to write a code looking like this: Then I stepped back a bit and thought: OnNavigatedTo is an event, and its return type can’t be changed from void. Therefore, is there a point in awaiting the last asynchronous call of the method, as this will provide no information at all to the caller? Would it result in a useless call to the dispatcher, just to execute an empty callback? Actually, is the compiler smart enough to detect this case and remove the seemingly useless await? Understanding what’s going on How to figure out whether the last await call has any impact? Let’s say I add some code atRead More →

By:
On:
Tagged: , , ,

That’s an issue happening from time to time. You need to do some lengthy computation on the UI thread (for instance, a page navigation), and you want to display a progress indicator to ease up the user waiting. It’s possible on WinRT thanks to the compositor thread: a dedicated thread for animations that don’t impact the page layout, making them run smoothly even when the UI thread is busy. Some built-in controls automatically use the compositor thread, such as the ProgressRing. So, what the issue could be? Letting the UI refresh Let’s say that when the user taps on a button, we want to displayRead More →

By:
On:
Tagged: ,

That’s a surprising behavior, brought to my attention by this question of Anthony Wieser on MSDN forums. Basically, he noticed that in the global exception handler, he could read the stacktrace of the exception only once. On the subsequent tries, the value would be null. The debugging part We’ll start by making a simple program to reproduce the issue. When the user click on a button, we throw a custom exception: In the App.xaml.cs file, we subscribe to the unhandled exception handler, and print twice the stacktrace of the exception:  Sure enough, the stacktrace is printed only once. The second time, the property is null. What’sRead More →

By:
On:
Tagged: , ,

In a mobile application, we often face the challenge of fitting a large picture in a limited, fixed-size space. One of the common solutions is to generate a thumbnail, but depending on the ratio of the picture the result will not necessarily meet the quality standards. In my case, I decided to try an alternative approach and mimic the built-in effect when pinning a picture as a secondary tile on Windows Phone, where the picture slowly scrolls from top to bottom (Ken Burns effect). There is no control on Windows Phone to produce this kind of effect, so I had to put together my own. FirstRead More →

By:
On:
Tagged: , , ,

Another new addition to the Windows 10 SDK is the SplitView. It’s used to display a side menu, such as the one usually called “hamburger menu”. It’s quite straightforward to use. The “Pane” property contains the code of the menu itself. The page content goes into the control itself. The “OpenPaneLength” sets the width of the menu. Lastly, the “PanePlacement” property indicates on which side of the page the menu will appear (currently limited to left and right, top and bottom doesn’t seem to be supported). The menu is opened by setting the “IsPaneOpen” property to true, and closed when the property is set toRead More →