What setup code should go in Form Constructors versus Form Load event?

Programmers that have worked with VB6 tend to put a lot of code in the Load event, in VB6 that event was used to initialize the form. But that’s not appropriate anymore in Windows Forms, the Form class can have a constructor. The .NET way is to initialize class objects in the constructor, there are very few compelling reason to not do so for the Form class.

The Load event runs right after the window handle for the form was created, just before it becomes visible to the user. You should only write code in the event handler that depends on having the handle created. There is not a lot of code that qualifies for this requirement except one kind: code that requires the window size and location to be known.

The design-time Size and Location property values of a Form are not the same as their actual values when the form runs on another machine. The form can get rescaled to accommodate the system font size or the video adapter DPI setting on the target machine. The user preferences play a role too, the user might have selected a different font size for the window caption. You don’t typically care about any of this, unless you want the window to have a particular position on the desktop or be aligned with some other window.

Writing code in the Load event that does things like initialize TreeView or ListView controls can actually dramatically slow down the startup time. When you do it in the constructor, Windows Forms doesn’t have to update the physical window yet, it hasn’t been created yet. Once the native control gets created, Winforms initializes it with a bulk update instead of one node/item at a time as will happen when the code runs in the Load event. Big difference.

Last but not least: you should never use the Load event, you should override the OnLoad() method. This ensures code runs in a predictable order when you (or somebody else) inherits from your Form class. IntelliSense helps you write this method, just type “protected onl” and press tab to have IntelliSense auto-complete the method. Note how you have a choice to put code before or after the base.OnLoad() call, that’s how you control who is the boss. You are the boss when you put it after, not often the correct choice btw.

Leave a Comment