I hear/read about it sometimes when talking about .NET, for example "managed code" and "unmanaged code" but I have no idea what they are and what are their differences. What are their difference, by definition? What are the consequences of using either of them? Does this distinction exist in .NET/Windows only?
Managed code is what Visual Basic .NET and C# compilers create. It runs on the CLR (Common Language Runtime), which, among other things, offers services like garbage collection, run-time type checking, and reference checking. So, think of it as, "My code is managed by the CLR."
Visual Basic and C# can only produce managed code, so, if you're writing an application in one of those languages you are writing an application managed by the CLR. If you are writing an application in Visual C++ .NET you can produce managed code if you like, but it's optional.
Unmanaged code compiles straight to machine code. So, by that definition all code compiled by traditional C/C++ compilers is 'unmanaged code'. Also, since it compiles to machine code and not an intermediate language it is non-portable.
No free memory management or anything else the CLR provides.
Since you cannot create unmanaged code with Visual Basic or C#, in Visual Studio all unmanaged code is written in C/C++.
Since Visual C++ can be compiled to either managed or unmanaged code it is possible to mix the two in the same application. This blurs the line between the two and complicates the definition, but it's worth mentioning just so you know that you can still have memory leaks if, for example, you're using a third party library with some badly written unmanaged code.
Here's an example I found by googling:
#using <mscorlib.dll>
using namespace System;
#include "stdio.h"
void ManagedFunction()
{
printf("Hello, I'm managed in this section\n");
}
#pragma unmanaged
UnmanagedFunction()
{
printf("Hello, I am unmanaged through the wonder of IJW!\n");
ManagedFunction();
}
#pragma managed
int main()
{
UnmanagedFunction();
return 0;
}