For a decade now, we've worked to understand the phylogenetic posterior distribution, as well as to develop non-MCMC approaches to inference. In this talk, I'll survey what we have learned, including our many mis-steps, and speculate about what the future may hold.