🚀 go-pugleaf

My $.02: 

My wife and I teach our boys (12, 9, 6) at home.  I tried
unsuccessfully a couple of times to teach the older boys something
about programming, first with Pascal and then with C.  This summer we
tried Object Logo, and the results have been spectacularly successful.

It seems to me that Logo succeeded where the others failed for three
main reasons.  First is the simplicity of Logo's syntax, second is the
immediate feedback of its interpreter, and third is the ease of doing
interesting graphics with Logo.  Adding current software-engineering
techniques to Logo would not, in my opinion, improve it in these
respects.

There are a couple of features of Object Logo that we have hardly
used at all, namely robotics and software objects.  In my mind
robotics is not conceptually all that different from turtle graphics.
There is a great difference perceptually, however, and I believe that
programming physical objects--things that are built and modified by
hand--will make Logo even more interesting and exciting to my kids.

Object Logo allows you to use traditionally "pure" Logo, if that's
what you want.  It offers in addition a systematic and complete
object-oriented version of Logo.  We have used OO facilities a little,
when they provided a natural way to deal with problems we were working
on.  As we attempt increasingly more complex projects, I expect that
we will increasingly adopt an OO approach.  This is partly due to the
fact that Object Logo requires the use of objects to do some things
(eg robotics and sound).  But it's also a natural way to deal with some
kinds of complexity.  So it seems to me that incorporating OOP into
Logo is a good thing to do, although it's likely to be useful only
after one has explored some of the traditional parts of Logo.

These remarks have had to do mainly with Logo as a programming
language, as opposed to Logo as a promoter of wide and rich
educational growth.  Traditional Logo offers a huge variety of
interesting ways to explore and experiment with a wide range of issues
in logic and math.  It gives practice and experience in analysis and
problem-solving.  As Pappert wrote, it can give kids their own
universe to fashion and control.

Other programming languages surely have the power to provide these
things also.  However, as my experience with Pascal and C proved to
me, other languages impose too many impediments; the language itself
must be mastered before its power can be used.  If some modernized
successor to Logo were to be planned, I hope that it would follow
Logo's lead.  Most importantly, the language itself must seem simple
and unobtrusive, but at least as full of possibilities as Logo is
already.


Graham Mark
Los Alamos National Laboratory
(505)667-8147