Path: news.media.mit.edu!bloom-beacon.mit.edu!usc!cs.utexas.edu!uunet!munnari.oz.au!yoyo.aarnet.edu.au!ntx.City.UniSA.edu.au!magill.magill.unisa.edu.au!kerrb
From: kerrb@magill.unisa.edu.au
Newsgroups: comp.lang.logo
Subject: Boxer answers
Message-ID: <1994Jan24.125223.103@magill.unisa.edu.au>
Date: 24 Jan 94 12:52:22 +1030
Organization: University of South Australia - Magill
Lines: 78

From:	IN%"disessa@media.mit.edu"  "Andy diSessa" 22-JAN-1994 04:07:01.28
To:	IN%"KERRB@Magill.Magill.UniSA.edu.au"
CC:	
Subj:	boxer

Feel free to share this.  If you need more info., write back.

--Andy
---------------------------------------------------
Precis on Boxer:

1) rationale for its development 

Boxer design came out of two perceive weaknesses of Logo.  (a)
comprehensibility -- including the well-known "humps" such as
variables for young children, procedures, list processing and writing
well-structured code.  (b) usefulness -- you just can't do enough that
is immediately useful in Logo.  This latter was actually more a
problem for teachers, as I perceived it, than students.  Kids LIKED
drawing with Logo, but teachers learned it "for their students," not
the best learning relationship.

Boxer has been a "principled design" development.  It's principles
of design, including the spatial metaphor (reusing spatial intuitions
everyone has to help people understand programmng) and "naive realism
(making a system that shows itself and allows full interaction
directly on the screen), have been widely discussed in publications.
We have worked in schools for about 5 years now, including elementary
school (a full-year course on physics), high school physics and math,
and high school biology.

The "grand image" is of a new literacy in communities of tool-makers
and sharers.  We've discovered collaboration is an especially strong
point with Boxer

2) Written resources -- available from where?

The best single published source is the two issues of the Journal of
Mathematical Behavior, 10 (1 and 2), 1991, devoted to Boxer. Also
diSessa, A. A. and Abelson, H.  (1986).  Boxer: A reconstructible
computational medium.  In a special issue on Teaching Programming,
Communications of the ACM, 29 (9), 859 - 868.  Also discussed briefly
in a recent article in CACM by Elliot Soloway.

An annotated bibliography of published papers and technical reports is
available from the address below.  (Send also paper requests.)  There
may be copying fees for papers ordered.

The Boxer Project
Attn: Eowyn Mader
EMST
Graduate School of Education
University of California
Berkeley, CA  94720


3) What sort of machine do you need to run it on?

We have an experimental version of Boxer available on the Macintosh.
It needs a 68030 or 040 Mac running System 7 and 32 bit addressing.
(E.g., SE/30 (minimally), LC III, CI, CX, VX, Quadra/Centris 610, 650,
700, etc.).  8 mb minimal, more realistically 12, plus hard disk.
Preferably at least 640 x 480 screen resolution.

The more complete and debugged Sun version is also avaiable for those
lucky enough to have workstations.

Both versions have demos and on-line documentation.

Berkeley is licensing these experimental versions (no guarantees!) at
$250 for single user ($500 for a group licence), though availability
is limited.  We are pursuing commercial development.  For information
and to place orders, contact:

Office of Technology Licensing
University of California, Berkeley
2150 Shattuck Ave.
Berkeley, CA  94704-1318