IsNothing versus Is Nothing

Does anyone here use VB.NET and have a strong preference for or against using IsNothing as opposed to Is Nothing (for example, If IsNothing(anObject) or If anObject Is Nothing...)? If so, why?

EDIT: If you think they're both equally acceptable, do you think it's best to pick one and stick with it, or is it OK to mix them?

Solution 1:

If you take a look at the MSIL as it's being executed you'll see that it doesn't compile down to the exact same code. When you use IsNothing() it actually makes a call to that method as opposed to just evaluating the expression.

The reason I would tend to lean towards using "Is Nothing" is when I'm negating it becomes "IsNot Nothing' rather than "Not IsNothing(object)" which I personally feel looks more readable.

Solution 2:

I find that Patrick Steele answered this question best on his blog: Avoiding IsNothing()

I did not copy any of his answer here, to ensure Patrick Steele get's credit for his post. But I do think if you're trying to decide whether to use Is Nothing or IsNothing you should read his post. I think you'll agree that Is Nothing is the best choice.

Edit - VoteCoffe's comment here

Partial article contents: After reviewing more code I found out another reason you should avoid this: It accepts value types! Obviously, since IsNothing() is a function that accepts an 'object', you can pass anything you want to it. If it's a value type, .NET will box it up into an object and pass it to IsNothing -- which will always return false on a boxed value! The VB.NET compiler will check the "Is Nothing" style syntax and won't compile if you attempt to do an "Is Nothing" on a value type. But the IsNothing() function compiles without complaints. -PSteele – VoteCoffee