[Edu-sig] seasonal challenge to calculator dominance in high schools

kirby urner kirby.urner at gmail.com
Sun Jun 23 16:37:33 EDT 2019


On Sun, Jun 23, 2019 at 12:57 PM Wes Turner <wes.turner at gmail.com> wrote:

>
>
> On Sunday, June 23, 2019, C. Cossé <ccosse at gmail.com> wrote:
>
>> I'll bet every one of those graphing calcs has also been replicated as a
>> phone app
>>
>
>> That's cool stuff there!  (yours)
>>
>
> Yeah, that's really cool. Was the MoCap (motion capture) done at the
> University of Nebraska Omaha (UNO)? They've a new Biomechanics facility
> next to the soccer field over there.
>
>
Thanks, I think the MoCap was done at a gait analysis lab, before facial
recognition got so good (actually it's not either / or).  They'd like to
know who that guy is over there on the camera based on his distinctive
walk.  However the Dr. Fuller First Person Physics initiative piggy-backed
on all that to study human motion as physics.  Turning on those contrails
was maybe my idea.

"First person physics" was my coin as far as Dr. Bob Fuller (no relation to
Bucky) was concerned, and he made me webmaster for the project.

> There are a bunch of cool videos demoing simulated agents learning to walk
> with evolutionary algorithms (mutation, crossover,  cost function)
> https://github.com/openai/mujoco-py/blob/master/README.md#usage-examples
>
> https://youtube.com/results?search_query=openai+learning+to+walk
>
>
Yes, my friend Gerald de Jong was a first adapter of "elastic interval
geometry" where every "rod" is a tension-compression spring governed by
mathematics.  He put creatures made as tensegrities in a simulation and
selected for which was able to walk furthest, of course adding a concept of
gravity + friction (traction).

https://youtu.be/_II-uESToOs  (done in Java, Gerald an early adopter since
JavaOne, when he came to visit me in Oregon on his way back)

That's the same Gerald I cite in my online Google slides, who came up with
a tetrahedron volume formula very like Euler's, but geared to give output
in tetravolumes. I have it in Python.


By comparison, my old offline graphing calculator is a frustrating piece of
> work with no QWERTY keyboard.
>
>
The issue is the personal workspace (PWS) and providing the ergonomics of
at least a cubicle to people that young.  Once you're college age, you have
a dorm with internet and library study carrels, we hope. But already you're
going into debt for that.

Babies don't know how to drive for Uber, if you know what I mean.

Like how will the younger kids ever get the political clout they'll need,
to get out of that "locker and tiny desk" rat race (eat lunch the nearby
fast food places, how convenient!).

How will they ever find time to develop a private stash of Jupyter
Notebooks in such crowded crushing daycare centers?  Might as well be jail.

At the School of Tomorrow, I assume students are kicked back in private
quarters, like in the movie (the kids each have their own room, not just
the dome kid).

In building a table out of a rolling cart, 2x6's, a melamine sheet, and
> some brackets, I had need for rigid body dynamics; to determine how much
> force would cause the table to fall over. After not finding any existing
> open source software with actual calculations and a few q&a questions with
> some equations and parameters, I considered trying to add support to
> FreeCAD (with cadquery and Jupyter Notebook) for rollover risk.
>
>

I had a gig sharing CAD online with middle schoolers.  Not Blender.
Something cloud.

Even the schools using Macs found it slow going.  We had Chromebooks as
backup.

What was it called again?  I'm scanning through a plethora of free online
tools and not finding it.

Maybe it was the wifi connection that was slow.  I don't think 3D CAD is
something you just learn in an hour.  The dashboard (GUI) is intimidating,
even though this is a "dumbed down" version.

We're all free to dream of our castle-in-the-sky curriculum (I certainly do
that a lot) wherein the hardware problem has already been solved, but short
of turning high rise floor space -- already used for cubicles -- into high
schools, I'm not sure how to solve the problem.  Making them use itty bitty
calculators or tiny computers is designing for claustrophobia.

One laptop per child (a goal, not an achievement yet), leads to a next
question:  where do you plug it in and get some quality time with it, and
the people reachable through it?  How do you keep it from being stolen
and/or banned in your state?  You're relatively little, surrounded by
bigger people who always think they know better.

Anyway, I think it's great we get to play with physics engines.

Even Codesters, which I use a lot to teach Python (just signed up for more
business), has primitive gravity and bounce effects you can toggle on and
off.

Not a full Python implementation (no yield statement, does include classes).

Kirby
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/edu-sig/attachments/20190623/c60f3e06/attachment.html>


More information about the Edu-sig mailing list