I've spent most of the weekend trying to grasp the quirks of 4 different poorly-documented programs that only run on Windows/Linux and the literature surrounding them so that I can figure out what of those has been implemented into 3 or 4 different R packages so that I can appropriately finish some analyses. At this point, I'm probably going to have to settle for something imperfect but 'good enough'.
I have 3 chapters that look like chapters. One of those still probably needs a day of work, but I can afford no more than that. I have 16 days to produce the next two chapters before I send my dissertation to my committee. I'm counting on those needing a lot of revision after the defense, but I hope Herb will find them defensible. They just have to be good enough.