Recently I had to implement a recaptcha in one of our projects. I was looking for an npm package, but even the popular ng-recaptcha package didn't support Angular version 18.
After checking the javascript code I came up with a simple solution, by building my own typescript service to wrap the javascript library from google.

One of my projects stores meta data of files in the database. The information is held in a FileEntity. Any other entity can then have a relation to this FileEntity. The idea is, that files are uploaded first and are referenced afterwards. For example when a file for a photo album is uploaded, it will first create a FileEntity and then a PhotoAlbumEntity that contains a reference to the FileEntity. The downside is that unreferenced files are using up space. So I had to find a way to find unreferenced file entities with Entity Framework Core.

One of our projects recently had severe performance issues which were caused by database queries. Eventually we figured out that the problem was the memory allocation caused by table columns which were all using NVARCHAR(MAX).

Today I had a task to consume a SOAP endpoint from an external source. The documentation says that base authentication is required but also a client certificate has to be used to establish the SSL connection. I was trying for several hours to accomplish both, but it seemed that it was not that simple.

You may have noticed in the last part, that whenever you leave the game, or even change the screen orientation, you will lose the current state of the game. In this part we are going to learn how we can keep the state of our running game.

In this part we are going to implement a scene. Scenes are helpful when you need to display different views in your game. Also scenes will make it possible to separate specific game code from the engine code. The first scene your game will show is usually a menu where the user can choose to start the game or enter the settings. Another scene might be the actual game where you render sprites that are controlled by user interaction.

In this part we are looking at important aspect of games, handling the user input. Without user interaction a game is not playable. Luckily it’s pretty simple to integrate user input in our project. In this article we will simply show or hide a sprite, whenever the screen is touched.

In this blog series I want to talk about a matter of heart. Since I was a child I always wanted to create my own computer game. I started with C++ and bought a book about how to create a game with DirectX. After a lot of time I managed to create my own games. But it was very complicated and took a lot of time. Four years ago I started developing a mobile game called “Mouse revenge”. Since I wanted it to become a mobile game I chose C# and Xamarin. In the beginning I was really struggling to find a good tutorial on how to create a platform independent game with Xamarin. So I tried a lot myself and finally found a structure that I could use to write only little platform dependent code and have most part of the game engine separated. After experiencing some bugs on specific devices I managed to have a stable version with zero crashes since months. I think it’s the right time to share my knowledge to everyone who wants to create his own mobile game with a good programming language.

Last week I have been reviewing some code of one of our newer projects. I noticed that for some projections, also Include() was used:

Recently a requirement for a project was that there has to be a registrations table with a column with the following format “2023_0001”. This format contains the year and a consecutive number. Each year, the number has to start with 1 again. In this post I want to present some possible solutions for this problem and weigh up arguments against each other to find a reliable solution.

Since Angular Material 15 the MatDialogs look quite different than they used to. Especially the font sizes are different. The CSS has changed:

One of the greatest features I have discovered in the last few years is expression expanding for entity framework core. Usually when you create a projection with linq you will use the Select() method and either use an anonymous type or map the values you need from the database to a class. These projections are often faster because they don’t have to retrieve all the columns from the database table. What I always missed was to be able to reuse a projection inside another projection. This can be helpful if you have expandable structures you want to provide to your API.

When Android 13 was introduced, I noticed in my Google Play Console that a lot of users are getting new exceptions almost daily. Mostly it were “java.lang.NoSuchFieldException” or “android.os.BadParcelableException” exceptions.

Recently during work, my team came up with the question how to query database entries that have a unique key. Should we use First or Single? We already knew that First will make a database call where it will limit the result to 1 row. Single will limit the result to 2 rows, to determine if there is more than one result and if so, can throw an exception.