Int32 or Integer/int?
Brad Abrams posted about the use BCL types vs. language specific type names (in sample code) today, which just happens to be a topic I’ve had to think about myself when writing sample code for my own articles, so I thought I would make a comment.
I’m with Morty on this one… I have the general rule that sample code should be written following the style and convention guidelines for the language. I’m a VB guy, so this usually takes the form of “Write VB code the VB way”, but there are exceptions…. For example, I use Integer in my Visual Basic .NET code, not Int32, but I like to use Int32 in Win32 API declarations; the specific size of integer is pretty important at that point so I like to make it clear… Beyond types it gets a little murky… I always use the System.IO classes for File access; VB.NET still has the Open/Put commands but I truly believe that those “legacy” statements are not as good/clear as the new System.IO classes… whereas in the case of Int32 vs. Integer (or MsgBox vs. MessageBox.Show()) it is merely a syntax/style difference.
I think using Int32 in a VB.NET/C# code sample makes it slightly less familar to a C++ developer (in the case of C#) or a VB6 developer (in the case of VB6)… and that makes the developer a little less comfortable, and therefore a little less likely to understand what they’ve been shown. The purpose of a sample is to explain the concept in terms of code. The purpose of doing in more than one language is to explain the concept in terms that the developer is already familiar with… well, following that same guideline and knowing that a VB.NET developer is more familar with Integer than Int32, they will benefit more from seeing the sample written using the VB data type names.
Of course, all of this can be ignored if Int32 was the class being demonstrated 🙂
Thoughts on this post? Feel free to reach out on Twitter!
2003-04-06 10:50 +0000
0b01fd9 @ 2019-02-21