[sugar] Seeking opinions on the viability of an idea

[Ian Bicking]

Mike Hunter wrote:
> Greetings programs,
> 
> I did a project in Graduate school four years ago that I think might be
> interesting for OLPC to look at, and I'm writing in hopes of getting
> some honest feedback from people about it.
> 
> My classmate and I wrote a program called "SymLogo" which implements a
> subset of the logo language with only graphical symbols:
> 
> http://www.mste.uiuc.edu/courses/ci407su02/students/mthunter/project-final
> 
> I recently moved to Atlanta for my wife to take a new job, and I am
> currently -- what's a polite way to say "unemployed...." -- weighing my
> options of projects to work on.  On the long drive from California, I
> read an article about OLPC which pushed me over the edge from "sounds
> fun" to "I should work on this!"
> 
> It's definitely a graduate school project and not a production tool at
> the moment, but my new-found "freedom" means that I could put some more
> effort into it.
> 
> During my working career I've been subscribed to several FreeBSD mailing
> lists, and it was there that I learned the following paradoxical course of
> these sorts of discussions:
> 
> "Hey everybody, I have an idea, if I make it would you commit it?"
> "Depends; show us the finished product and we'll talk"
> "But I'm not going to do the work unless I know you'll use it"
> ...
> 
> But when you stay on the lists for long enough, you learn that this is
> actually the correct approach, because there's really no other way.
> 
> Having said that, I'd appreciate any initial guidance the group has on
> this.  Is this idea already covered by a different application (I've
> been reading up on eToys; very cool stuff, but I think there's room for
> this, especially for younger kids and for the no-need-to-localize
> factor).

It does have a lot of similarity to EToys, which also has drag-and-drop 
programming that uses a fairly minimal (and localizable) set of 
instructions with graphical hints and whatnot.  In theory you could 
develop a purely graphical subset of that scripting language, which 
would run in EToys.  I can't speak to how easy that would be.

There are also a number of graphical systems under the general moniker 
of Karel the Robot.  I can't even tell which one is the original 
anymore, and they have a wide variety of syntaxes, most of which would 
translate into icons fairly well.  They are usually more constrained 
than turtle graphics, and the languages are much more constrained than 
Logo.

My own opinion is that an system like this will be useful mostly based 
on how entertaining/useful/educational the possible activities are.

There's several different platform choices you could make as well, but I 
think the activities are a more important concern, and would drive a lot 
of the other choices.

-- 
Ian Bicking | ianb at colorstudy.com | http://blog.ianbicking.org