[Edu-sig] The trackball reality

[Jim Harrison]

on 10/24/03 12:59 PM, Arthur at ajs at optonline.net wrote:

> "Learning" needing to be in quotes, because that is the crux of the matter.
> Because if one defines "learning" as what one might gain from interfacing a
> computer with a trackball, then one will conclude that the trackball is a
> great interface for learning.  Of course - its a simple tautology.  And if I
> prefer to define "learning" as what one gains by interfacing with the
> keyboard  - I simply can't loose to trackball.

I'm not entirely sure of the point here, but it seems a truism that if one
is trying to teach use of the command line, a mouse/trackball interface is
not the way to do it. Vice versa also applies. And if one is trying to teach
material that has nothing to do with computing, the interface should either
get out of the way so that students can deal most directly with the material
or if appropriate the interface should reinforce some aspect of the
learning.

As an alternative view, the programming language that I found most
productive in the past and that made me understand OOP was Prograph.
Prograph is a completely graphical, iconic language/IDE that is compiled
directly from diagrams with no text intermediate. Nodes in the diagrams
represented program actions at approximately the same granularity as Python
statements. There were no variables in the usual sense and program execution
occurred as data flowed from node to node across connecting lines (this was
represented graphically during execution). I don't think high level textual
languages represent what's happening in the computer any better than
Prograph did, and moving from Prograph to textual OO languages didn't
present a conceptual problem. A mouse/trackball was the right tool to use
with Prograph whereas a keyboard is the right tool for textual languages.

Jim Harrison
Univ. of Pittsburgh