[Edu-sig] creating an interface vs. using one (Michel Paul)
kirby.urner at gmail.com
Sun Sep 24 21:55:09 CEST 2006
On 9/24/06, Arthur <ajsiegel at optonline.net> wrote:
> In both our battles - yours against the TI calculator, and mine against
> the 3d Ninja Turtles - VPython-like functionality is essential. It
> demostrates to your Problems that we are working with something with
> out-of-the-box more power than what they are committed to, and
> demostrates to my Problems that, being committed to working with and
> teaching about abstraction, we are progressive in the tools we use to do
> so - could be doing EA huminoid avatars and anthromorphic turtles if we
> chose but choose not, because they are distractions in addressing,
> teaching and exploring the world of abstraction.
> Making both those statements by the inclusion of vpython-like
> functionality in the standard distribution is to me a business decision,
> and a good one.
We simply disagree on strategy, and not by much. You focus on the
python.org download or core + standard library as "the distro" whereas
I'm thinking more in terms of a complete Linux or other operating
system bundle (Microsoft not precluded from distros -- that's what the
vendors do, e.g. HP includes Python as a part of its default XP
distro, on Pavilions and such).
I'm happy to have VPython in the distro (mix), I just don't think
bloating the python.org install file is the way to go. Get a whole
computer already outfitted, or rely on school IS (information
services) to burn a Gold Disk locally, with everything the faculty has
agreed is a keeper. Plus faculty gets to go out on the Internet for
more experimental stuff.
Vpython is not languishing in obscurity because it's not in the
default installer. The barriers are more cultural than technical.
Lots of schools have broadband and could be running Pygeo tomorrow if
the admins were OK with it. But they're not, and not because of
anything to do with Euclid or Klein, both already safely defanged and
property of the Ivory Tower.
But as I've said before, the whole culture of open source software is
still eyed with suspicion, in part because local biz moguls have done
their darnedest to make sure it's perceived as Communist or something
else scary. I've fought that by reminding people that Snakes on a
Plane are inherently even scarier yet i.e. we're capable of something
more home grown (i.e. let's stop blaming the Russians for kwel). But
that strategy only flies with some audiences.
Anyway, I think I should come up with a PEP proposing we rename
Standard Library to Forlorn Little Graveyard. Then people might
refocus and stop treating it like a Hall of Fame. If you wanna be
famous in Python Nation, don't suppose inclusion in the Standard
Library is your last/only chance. Agitate for inclusion on a distro,
get your work linked, noticed, included. Give use the screencasts
about it (e.g. on youtube or video.google).
In sum, whereas I think Pygeo has many distribution options, I think
trying to bloat the default installer with Vpython is *not* the most
efficacious route to that end. You'd do better to team with
like-minded math profs on your end, and do an Arthur Siegel Distro, or
call it something else. Branding is an art form and I won't pretend
to tell you how to run your business from this long distance.
More information about the Edu-sig