RE: [Aimmath-developers] changes

[Ken M. <mo...@pt...>]

>   I may generate a lof of disagreemet with what I am going to say but :
>
> > I tend to agree with Ken. There seem to be a lot of packages that were
> > new in Maple 7, that make coding of AIM questions slicker and easier.
>
> Although I agree, I would propose that the extra cool, slicker and easier
> tools that Maple supports
> are used in a sensible way and probably making sure that other CAS support
> them (maybe not exactly the same code but but at least to be possible to
do
> something similar - I have no idea what you are referring to anyway) so as
> we can easily re-use questions and possibly code in case we will not be
able
> to use Maple after some point or in general to accomodate other projects
> (like Chris' one in Birmingham or our one ghere) that look into
> using freely available CAS.

I think one of the main advantages of AIM over similar math assessment
programs like WeBWorK, and in fact even over MapleTA, is that we can write
our questions with the full power of Maple at our disposal instead of being
limited to an ordinary language (e.g. WeBWorK writes their question in
Perl).  If we try to restrict ourselves to using only the features of Maple
that appear in ordinary languages (loops, if then, assignments, etc), and
perhaps a small subset of the CAS features that are available, I feel we are
then moving in the wrong direction... back towards a system like
WeBWorK/Perl.

I do not think it would be practical to translate most of the AIM questions
I have written from Maple to another CAS language if that option ever
becomes available, whether or not we use the extra tools that are available
in the newer versions of Maple.  However, for the question author, the more
slick advanced tools we have at our disposal, the easier it is to write,
maintain, and debug the questions themselves.  So I wouldn't want us to
cripple the system or limit what we can code in our questions in order to
try to make it compatible with everything.

regards,
KEN