Question

I am creating a logger for an application. I am using a third party logger library. In which logger is implemented as singleton.

I extended that logger class because I want to add some more static functions. In these static functions I internally use the instance (which is single) of Logger(which i inherited).

I neither creates instance of MyLogger nor re-implemented the getInstance() method of super class.
But I am still getting warnings like destructor of MyLogger can not be created as parent class (Loggger) destructor is not accessible.

I want to know,
I am I doing something wrong?
Inheriting the singleton is wrong or should be avoided??

Answer 1

Leaving the merits of singleton pattern aside (there is a school of thought that describes it as an anti-pattern) it should not be necessary to subclass it to simply add static functionality. Here are language-specific approaches that I would prefer to subclassing a singleton:

• Use a free-standing function enclosed in a C++ namespace
• Use an extension class in C#
• Use a helper class in Java
• Use a Category in Objective-C