Duplicated code is an inevitable property of every software system---including yours. The good news is that not every clone is a threat to quality. The bad news is that some clones may already have caused problems or may facilitate future problems. To separate the good from the bad, we have to analyse the evolution of clones which provides rich information about the nature of individual clones. In this talk, I present our incremental clone detection algorithm that detects and tracks clones across consecutive versions of source code. The algorithm can be used for retrospective analysis of a program's history as well as live analysis during development. The resulting clone evolution model provides information on which of your clones you can consider friends and which of your clones you should distrust.