Bring out the bossy

[infryq]

Yesterday a couple things went awry with the aggregations, but I managed to clean them up and submit in time for the automated pickup. I did a pile of renames until we finally decided to maybe stop and think about what the names should be and get buy-in from all stakeholders on them. R revealed that he either had not read, read but had not retained, or had some kind of mental block about all my reports for the last two weeks, since it turned out the one new sensor we’d developed that wasn’t a raw import from another institution, wasn’t actually supposed to be going up in the interface at all this week. I had a good talk with him about all the reasons why that’s a problem. He agreed, and scheduled a larger talk with the indicator leads and RR about how to fix it going forward.

I went to a meeting with F and the team from M about how our smother works, and also got some reassurances that csv files from international browser locales don’t break the R scripts we’re currently using for analysis. I tested out the new standard error code and it worked beautifully. Lunch, then I attempted to merge in L’s branch for automated integrity checks. There were too many diffs, which is odd since it should have been an exact match to what’s in the API. Figuring that out later. Did a retrospective with the indicators team, which largely turned into a mutual love-fest, my second favorite kind of feedback session. Team leads meeting was productive; we hashed out ideas for written specs and making rigorous use of the github issue tracker that should help prevent things from getting so persistently lost in future. Helped A with some documents for kicking that off at today’s group meeting, then called it done.

I finished Exit Strategy and started Network Effect, which is full length. Good.

Dinner, a very stubborn crossword, DS9, snack, sleep.