[Edu-sig] Shuttleworth Summit

kirby urner kirby.urner at gmail.com
Fri Apr 21 15:55:05 CEST 2006

Some thoughts:

* wx would not be appropriate as a PySqueak substrate, as it's too
oriented around our existing mental model of widgets (pulldown, popup,
canvas, button), whereas Squeak's more generic concept of graphical
object defies most widespread OS conventions.  Plus wx is fragile.

* IronPython is a likely implementation candidate for any such
environment, because it'll have that Pythonic glue language role in
.NET/Mono.  Any graphics intensive engine we're talking about is
likely the target for several languages (development costs would
justify such a cosmopolitan approach, plus lift a lot of weight off
PSF).  Learn it with Python on Mono, then find another language to
fall in love with.  It's OK, and it's still the same CLR in any case
and so in that sense remains familiar.

* Squeak itself is already very usable, so in terms of the
Shuttleworth trajectory, I refuse to see a bottleneck i.e. refuse to
imagine *any* new Frankensteinish Py* is so on the Critical Path that
all curriculum writing must cease and desist, while we wait for the
Software Gurus to write it.  On the contrary, we're not really waiting
for Software Gurus period, although we admire their work and credit
them for making our project possible (now, already, not "when it's
done (whatever 'it' is)").


More information about the Edu-sig mailing list