On 12/17/05, kirby urner <kirby.urner@gmail.com> wrote:

These "Prosperos Books" (some from O'Reilly?), could be used in a community college setting, sometimes in testing situations i.e. under controlled conditions that measure student performance in some pre-agreed-upon way.

For example, under the heading of "useless Python" we have:

...

...

N=100 sum([i+j for i,j in zip(range(1,N+1), range(N,0,-1))])/2 5050

OK, lots of fun syntax, but deployed in a surreal fashion i.e. no one would want to really write a function that way right? Surreal is a key word here. I think we've learned from the Head First experiments, among others, that humans crave bandwidth. One reason kids tune out, get bored, and therefore perform poorly, appear stupid, fall behind, is they're busy day dreaming. Anything to compensate for the tedium of low bandwidth classroom pedagogy. There's a reason we call it "pedantic" (the pace of foot soldiers -- trudge trudge trudge). What I'd like to provide is high bandwidth, over-the-wire learning experiences, preferably on big plasma, with sensurround sound, that fully leverage our capacity for imagination. Purists will whine about all the "irrelevant noise" stuffed into our pipelines, but perhaps they're just not attuned to all the cultural allusions, sidebar tributes, cameo appearances. Like, the above Python fragment ain't so useless in a dynamic worksheet context, wherein we have an embedded movie re-enactment of Baby Gauss, chewing on 1+2+3... 100 i.e. the 100th triangular number, and deriving 101 + 101 + ... 101 (100 times), with thought balloons showing "as if" he were thinking in Python (two ranges, one reversed, zipped to pairs, then pairwise, then globally summed -- then divide by two to elimate the doubling). And for the Nth triangular number:

...

...

def tri(n): return sum([i+j for i,j in zip(range(1,n+1), range(n,0,-1))])/2

...

...

tri(9000) 40504500

Which is of course completely ridiculous as we've already shown we don't need the memory-intensive range bit:

...

...

(9001*9000)//2 40504500

...

...

def tri(n): return n*(n+1)//2

Triangular number: the number of one-on-one handshakes or dialogs betwixt N participants, this time using N(N-1)//2 i.e. every "me" (N of us) times "everyone who ain't me" (N-1 others) divided by 2 (because one handshake works both ways as a greeting). Dynamic workbook: phones ringing, fragments of dialog. Consecutive triangles pile as tetrahedra, leading to more animations and more fill-in-the-blanks (using tri, already defined, define the Python function tetra). You could have real human teachers, even in real time, doing some of the teleprompting. A dynamic book doesn't presuppose everything's canned. There's room for dynamism even of a noncomputable sort, given the human element. I think that early Python Love Story out of Yorktown HS is prototypical: real human actors engaged in role playing, interspliced with dryer fare. With millions of kids growing up on Hogwarts etc., the demand for something a bit more engaging than Everyday Math ("Jimmy buys six eggs at .39 apiece, if he breaks two coming home..."). We're doing our kids a disservice if we plan around these dry-as-bones text books as our only mode of delivery. I'm not against text books per se. I *am* against text books as the only option. I think the publishers with compassion will pioneer the newer options. From what I've learned at OSCONs, we have every reason to expect more from XUL etc. There's this instinct to force junior to buckle down and learn hard stuff the hard way, i.e. it can't be serious if it's not dry-as-bones, anything glitzy is pandering/spoiling. I question that instinct, distrust it. It smacks too much of "we never had that much fun as kids, why should you?" We're not just teaching "mathematics" or "Python" or "programming" in a narrow sense. We're showing off what multimedia might do for us, encouraging students to emulate and then improve upon the workbooks themselves. We're teaching art, not just science, and maybe in that order for a welcome change. Kirby