Today I finished a working version of the new dynamic SN search scheduling infrastructure, which Patrick can hopefully just hook up to the existing scheduler, thereby enabling newer and smarter observations. No real fancy night-wide optimization, but at least it is smart enough to pick, at each moment it is asked, which field has the lowest cost function given what it’s done so far, and that’s probably good enough for the time being. If you’ve heard that 2/3 of the time taken in developing software is in testing and documentation rather than just writing code, you heard right.
In other news, I cycled up the mountain again today for the first time since May! Just in time to catch all the wattles in bloom. Looking forward to coming out of winter hibernation.