Encoding of process output. In the case of returnStdout, applies to the return value of this step; otherwise, or always for standard error, controls how text is copied to the build log. If unspecified, uses the system default encoding of the node on which the step is run. If there is any expectation that process output might include non-ASCII characters, it is best to specify the encoding explicitly. For example, if you have specific knowledge that a given process is going to be producing UTF-8 yet will be running on a node with a different system encoding (typically Windows, since every Linux distribution has defaulted to UTF-8 for a long time), you can ensure correct output by specifying: encoding: 'UTF-8'