I've returned to an old theme for me: tabular recordkeeping is at the core of civilization (silly-vization), and today that means getting steeped in databases, not just Excel. Those who lose touch with tabular bookkeeping, drop out of the middle class, according to this theory.
Whereas I consider a school "elite" for offering SQL to middle and high schoolers, that's not the right standard. That's like saying HTML, English grammar, parts of speech, times tables, are all dispensable, now that we have cartoons. I may have felt that way as a kid, as I adored cartoons and HTML was not yet invented, but looking back, I see "passing the torch" as a matter of relevant skills sharing, not only, but including.
The compromise, for me, would be more didactic cartoons, puppets and so on. Big Bird never taught us about complex numbers and none of the muppets shares SQL. Why? Sometimes it seems the more vital the knowledge, the drier the bones. Scare away the most imaginative and you'll have your ideal labor pool of the most compliant, seems to be the filtering mechanism.
Our sandbox is for learning SQL is of course SQLite3, given Python as sqlite3 built right in.
This isn't just a practice space, but a place to maintain legitimate files. The personal workspace of today and tomorrow has long included an electronic file tree, and those forced to adulthood with now exposure to filesystem trees have been cast by the caste system as people without office skills. We call it the Digital Divide.
You won't get to hold office if you have no office skills (e.g. SQL). The school system helps separate the two groups, often without letting on that this is what it is doing.
Lists of airports, favorite websites, stamps, coins... the literature is packed with examples, more or less practical. SQL is a whole separate language from Python, and seeing them both happening at once, in a single script, is most instructive.
On top of all this promotion of SQL in grade school, I put spatial geometry as a topic. The polyhedrons, in particular, need to be filed away in various ways, and retrieved accordingly. The inventory each has, of corners, faces, edges (V + F = E + 2) is an obvious starting place, followed by tags alluding to category, such as Archimedean, Platonic.
Do we endeavor to include Volume as a column? Given an edge length, why not? In fact, we might even carry two volume columns, as our "elite" curriculum is fluent around tetravolumes as well as XYZ volumes. In fact, those Jupyter Notebook examples are done, open source, available.
