I finished those online courses on SQL. There was a bit of cognitive drift into “other data models” (i.e., JSON and XML) towards the end, but all in all it was a positive experience: I really did learn a lot, and more important, I’m now very comfortable using these new things I learned.
(The XML foray was an interesting thing in its own right: these courses — and this is probably true of most “free courses” on the internet — are about 10 years old, which translates to “about the time XML’s heyday was starting to fade.” The course touched on XQuery and XSLT, but there was a definite “we suspect you won’t need to know this in the future” vibe about the lessons, and here in the future I had a hard time even finding a way to run the demos, and much of my internet research consisted of reading articles with names like “Why Should We Care About XML Anymore?” I eventually resorted to a bash script — brutishly practical, my workhorse go-to — to invoke the bloated, oh-so-elegant Java classes I found online to perform the queries.)
Anyway, I was pretty impressed with myself, both for sticking it out and for actually learning something from these courses. I think that the “internet course” format is something I take to pretty well, at least the ones I found at edX, and in fact something I enjoy spending my time with. Naturally enough, I decided to continue by taking another set of courses, and this time I’m giving R another go. I’m about halfway through the third course, in a series on data analysis put out by Harvard, currently working through visualization (graphs) and ggplot, and once again I’m doing well and loving it — so far. Only time will tell.