I am in Salt Lake City for the SDSS-III BOSS meeting. On the first day, there were a series of talks about the hardware state of the survey, and whether or not the early data (and the things they imply about hardware and targeting) are meeting requirements, as set down in a very detailed requirements document. Highlights include David Schlegel (LBNL) explaining why you should fit redshifts to the individual-camera exposures and not to any co-added spectrum, Adam Myers (UIUC) going through great detail about how complicated and uncertain the quasar target selection is, and Bill Carithers (LBNL) noting that for many science frames we don't (for reasons I didn't quite get) have arc lines shortward of 4000 Angstroms, where we care deeply about the wavelength solution for the quasar-absorption-line density field. It was not gloomy, though: The survey is underway and the data look excellent. Almost all non-spec items had paths to spec and I anticipate an amazing data set.