dybr

 Ok, deployed the diagnostics update for privacy job to prod. Ok. Half an hour later saw those errors properly reported. Posted it on slack.

And then kaboom, Chris comments: "oh, these are just test data, and the errors are legit." And Karen added: "yes, I have to find time to clean up the privacy database on prod". И эти люди запрещают ковырять в носу. Chris pretended to be a big proponent of quality and reliability, and now what.

Well, whatever, I did my part. And I'm not on pager duty anymore. If I were, I'd have created a jira case.

Karen is still reviewing my pr, the one that Chris refused to review. She asked me to write a description of the changes. That's about 6 lines of code in business logic, and about 200-300 lines of tests and test data. Yes, I added the description, it's important, is not it. Still not approved.

A good news is that we now have Newrelic. Which is way way more reasonable than Kibana. I just took a look, found some errors reported, investigated, cool. If I were on PD, I'd have created a jira case.

Standup, and again I had to explain the status. The status is: our pardners are totally happy, and the only thing left is to produce as output, on demand, exactly the same data they send us. We did not, before. We had an idempotent monad, so to say. 

Also, played with converting one project to use scala 2.12 (no shit). Turned out that somewhere deep, in the depths of sbt and ivy2, sbt looks for jars in e.g. ~/.ivy2/localcom.typelevel - neat, right? Asked people around, nobody seems to know. (sigh)