Referential Transparency

问题:

What is the meaning of the term "Non-observable" when used in context with the term "referentially transparent" in functional programming?

回答1:

As you might know, the term "referentially transparent" means that the value of expression can depend only on the values of its parts, and not on any other facts about them.

For example, it cannot depend on the following:

• Whether some part of expression is already evaluated or not (in a lazy language)
• Whether two equal values are shared (are pointers to the same location in memory) or not
• Whether a data structure is cyclic (i.e. its pointers create a cycle) or not

All those facts about the current state of the program are either true or false, but no expression can change its value depending on them. So those things are called non-observable.

This webcomic and its discussion on reddit might enlighten you as well.