[Axiom-math] feedback

[root]

*,

MIT has a lot of courses online. In particular, it has a fair number
of math and science courses online. (http://ocw.mit.edu)

The discussion at work today (I'm at City College of NY) is a proposal
to develop a common-format targeted front-end to Axiom which would
match and support the MIT math and science online courses. Of course
we don't have the man-power or expertise to develop such a large range
of courses in-house. Nevertheless, I think the idea has a lot of merit
for the education community and, in particular, the Axiom community.

Clearly the MIT math courses are reasonable for implementation in
Axiom (for example, Strang's Linear Algebra). And the Linear Algebra
course is clearly the basis for the Mechanical Engineering
course. Both of these are probably well addressed by the 3 Ms
(Mathematica, Maple, Matlab).  I don't believe, however, that there is
a coordinated body of code for the range of available courses in any
of these systems although pieces exist for some of them.  Axiom is in
a "come from behind" position at the moment.

I believe that Axiom's new open source nature gives it a major
advantage over the 3Ms since both professors and students can read,
modify, and create new algebra which could be freely shared worldwide.
Since we're trying to get Axiom shipped with all of the available
operating systems (Linux, Windows, Macs, etc.) it will be available on
all university and student desktops worldwide. This will give Axiom a
common foothold in the education market.  A common front-end approach
would lower the learning curve for students so less class time would
be spent teaching students to use a computer algebra system.  A
connection to the online courseware would make it very useful.  And it
would give widely available tutorials for Axiom so everyone benefits.

Is it reasonable for an amorphous, worldwide community to try to do
coordinated development of software around a single MIT target? Would it
be reasonable to convince, or hope to convince, MIT to host the same
courses in multiple languages? Could we figure out a way to develop a
grant/subgrant structure that would attract developers for particular
courses? Is it possible that any granting organization would fund
developers worldwide? Is it politically possible within Universities
to expect that MIT's presentation of Linear Algebra would supercede
locally developed courses? If not, is it reasonable to design the
software to allow selective order of topic introduction? Can we design
courseware that would handle, say, the top 10 selling Linear Algebra
books? Will professors allow students to use Axiom on tests (my math
professors would not let us bring calculators to class, including my
Advanced Calculus class :-) ). Can we arrange the issues and try to
address each of them in some coordinated fashion?

It is clear that free and open source development can have as large
an impact on education as Linux has had on the operating system area
if we can build a community around the effort. One of the major issues
is that building software gets little or no academic credit and even
less grant funding. Perhaps a coordinated effort can change both of
these mindsets.

If you have a few minutes I'd like some feedback on this idea.
Please reply on the address@hidden mailing list.

Tim Daly
address@hidden
address@hidden