LISA14 - Day One
This morning I jumped right into training with "Statistics for Ops: Making Sense Out of Data", taught by conference chair Nicole Forsgren Velasquez and Kyrre Begnum. The class started off with this warning: "If you thought stats class was a good place to take a nap..." Point taken, but no danger of that -- this was an awesome course, and my only regret was that it wasn't all-day. Both instructors were engaging and funny, and the material was perfect: working in pairs, and looking at 6 months of data, we explored what it means to purchase servers to cover 50%, 95% or 100% of your expected users. The results were not what I expected, and the morning was full of "OMG, NO WAY" moments for me.
During the lunch break I got to talk to someone who was at their second LISA; the first one was in 1996. The feel was the same, though: subject matter experts, and easy conversations with them. Of course, the difference was scale...but no matter what scale you're at (and no matter the decade!), there's a lot to learn from people about new ways to get your job done. Sometimes it's good to hear that things are exactly the same.
In the afternoon I went to the R for Sysadmins course, which dove into using the R programming language for statistical analysis. In some ways this was a continuation of the morning course (not least because our esteemed conference chair was one of the instructors here, too), but it was also a dive into a rich language with a lot of packages for tackling different things. The focus was less on "You have to do X by the end of the hour. How will you do that?" and more on "Here are the three different ways you might tackle that problem" -- but it was good, nonetheless.
The evening was spent discussing (by which I mean "me asking questions, then listening and taking lots of notes") the challenges of our profession with other folks: what do you do (indeed, what can you do) when poor upper management is taking down the company you work for? What say do you get in what other programmers do, or are obliged to do, when you're not a programmer yourself? How do you remain engaged in an important discussion that you've seen N times before over the last twenty years? In the end I did not make it to the games room as originally planned -- but I did meet fellow blogger Mark Lamourine, workshop leader Cory Lueninghoener, and Hadoop Tutorial Wonder Leader Jennifer Davis in person for the first time, which was an awesome thing.