Our engineering day was derailed partially by a great talk by Nathan Smith (Berkeley) about peculiarly luminous and ultra-luminous core-collapse supernovae. This talk was so beautifully structured, so clear, and so full of remarkable and tantalizing results, that my engineering team (read: Lang and Mierle) both wanted to quit astrometry and start working on extremely massive stars. They insisted on going to lunch with Smith; we spent lunch time talking about fundamental observables in astronomy, many of which were touched on in Smith's talk: He could measure delay times to explosive events with kinematic measurements, distances to objects by relating angular kinetic times to linear kinetic times, radiative and gaseous properties of nebulae by measuring families of emission lines, and three-dimensional structures of expanding nebulae with highly featured Balmer emission-line profiles. What a cornucopia of astronomy! Mierle, of course, suggested that we automate all of Smith's data analysis. Smith agreed that in any LSST-like future, automated spectral analysis will be a requirement.