This is incredibly impressive! I'm curious about the technical aspects. What were the biggest challenges in using Gource to create this visualization, especially with the large number of repositories?