I've been graced with a new VFP 9.0 Professional, completely legit, thanks to connections and UPS Ground, and last night polished off a demo screencast of inhouse academic assets, stuff we use to harvest clinical data for outcomes research, in support of the hospital's teaching functions ("what works for how long" is, after all, critical clinical information).
Meanwhile, on edu-sig, we're on the subject of teaching Python to non-CS majors, people with only a casual need to code, as a part of some humanities major for example.
"Do we share about OOP (our so-called object oriented paradigm) with the tourists?" might be a good way of capturing a key question.
I say that we should, but leveraging recog over recall skills, i.e. we make it look easy, with cartoons and such, without putting lots of pressure on students to cough up or regurgitate, i.e. you don't have to try this at home (but feel free if you want to, not a dangerous activity (though frustrating and tedious at times)).
I still have a high opinion of Visual FoxPro as a development environment by the way. Xbase was never a stupid language (though it may be used stupidly), with its procedural pointers to tabular data, combined with fast relational indexing, exposed in control structures. Best of all, it has an interactive command line (the infamous "dot prompt" in dBase), like Python does.
Then came the OO revolution, with Microsoft buying Fox as a dBase killer, once Borland's plans, to stick with the brand "dBase" (Ashton-Tate's) became clear.
Did the language die an ugly death at Microsoft's hands? No, I don't think so. The product remains an elegant solution in many situations and I'm glad Redmond keeps putting muscle behind it, despite end users going gaga for Microsoft Access (a part of Office no less). VFP wasn't just about using objects from other languages (more the VB model), had its own class defining syntax, more like in "VB#" (i.e. VB .NET).
What's great about xBase is you have this procedural way of stepping through tables and you have embedded SQL.
Sure, I can see why we'd use Oracle, Postgres or MySQL as a back end, but when it comes to creating a usable user interface, I still think the "thick client" has a future.
Making everything work in a browser "just because we can" isn't a deep enough philosophy to persuade me otherwise -- not yet anyway, XUL notwithstanding.