Comparison and versioning of scientific workflows

Abstract

Scientific Workflow Management Systems have become a widely used tool to orchestrate a sequence of programs, defined by scientific workflows, to build complex computer simulations. With the aid of these workflows, scientists develop their models for in-silico experiments. Although these workflows are very dynamic and need a large number of changes in their definition, very little effort has been done to support configuration management of these definitions. This work presents a configuration management model developed for scientific workflows, which includes both version control and diff/merge algorithms.