One of the earliest object oriented programming concepts I had trouble with for long time. The anomaly arises from the naming: I couldn’t associate the names mutable and immutable to something reasonable. Sometimes it seemed they relate to mutuality, sometimes to muting, and so on. Moreover, briefly I often confused mutable with mutex. Later came to realize they have rather more to do with mutation. Yes, mutation from biology. Wikipedia says
In biology, a mutation is a permanent change of the nucleotide sequence of the genome of an organism, virus, or extrachromosomal DNA or other genetic elements.
In the same manner, in OOP, mutable objects are on which mutation or modification is possible and immutable objects—on the other hand—are on which mutation is not posssible. To put more simply, attributes of a mutable object may change after its creation, but an immutable object’s attributes cannot change once it is created. This mutation connection makes meaning of the terminology more obvious.
A very obvious example of immutable and mutable classes are the String and StringBuilder(also called mutable string) classes respectively in Java, C#, and some other languages.
An assumption on an immutable object remains true throughout its lifetime; a very crucial property in many situations. Immutable objects are intrinsically thread-safe.
And the downside is that you may have to sacrifice performance for immutability: difference between the String and StringBuilder concatenation methods illustrate the fact.
External Links
- Mutable and Immutable Objects tutorial at Java Ranch by David O’Meara.
- C# StringBuilder source code at Microsoft Reference Source.