Transition documents

[infryq]

Yesterday I converted the team leads notes into announcements for the all-hands meeting, dropped the demo of the 14-day direction computations on a server so it’s easy to link to, and got stuck down a rabbit hole of converting the R notebook to make it easier to compare N direction calculations directly. Went to the big group meeting, asked about relative risk with an individually-tested but not fully-tested API for the data versioning system that was set to be deployed today, and ultimately decided to try and test it in the staging rig first. It was a good idea; we found a few problems that the Docker tests don’t look for, so we’ll be able to fix those today instead of getting an unpleasant surprise in production. Met with T and discussed hiring my replacement; he’s of course sorry to see me go but happy for the new direction my career is taking, which was nice to hear. Spent the rest of the afternoon rebuilding the automation for the surveys sensor to use the new codebase.

Dinner, crossword, noodled around on Twitter for an hour, then snack, measured ingredients for another batch of nut and seed bars, and sleep. Dreamt of a bleak but hopeful sortof novel set in nineteenth century England, solving mazes and giant mechanical architectural puzzles.