From braida@telly.mit.edu Mon Mar 13 08:20:16 1989
Date: Mon, 13 Mar 89 07:26 EST
From: braida@telly.mit.edu
Subject: Data Abstraction
To: hal@zurich

Having spent some time watching students work PS5 in the lab, I now
realize that the reason use of data-abstraction techniques is so hard
to learn is that we are not consistent about using it ourselves.
E. g.: (define p1 (make-polynomial 'x '((2 1) (0 1)))) instead of
(define p1 (make-polynomial (variable 'x) (adjoin-terms (make-term ...)))).
Also, if this problem set is given again, it might be a good idea
to have an exercise on sorting a termlist. The term manipulation
algorithms don't work if the termlist is '((0 1) (2 1)) instead of
'((0 1) (2 1)). A student claimed that he wasted three hours
as a result of not appreciating the sorting problem. Also, E 2.63 is
not as precise as it could be about what /poly should return. This
is a memo for file, rather than a complaint (after all, I constructed the PS).
I think the exercise was a good one, and hope it will be modified and used
again in the future.


