Report on CoMath `95 Maple V Conference
James Friend
August 17, 1995
- INTRODUCTION
- BOOKS AVAILABLE FROM SPRINGER-VERLAG RELATED TO MAPLE V
- FEATURES IN THE NEW VERSION OF MAPLE V: REVISION 4
- SUGGESTIONS AND PROBLEM DISCUSSION
- NEW INFORMATION ON MAPLE
While the conference was not what I had expected, it was useful.
Approximately eighty people were in attendance, and many of them were math
educators at either the college or high school level. A number of physicists
and mathematicians attended, and a few engineering students were there as well.
Held on the San Joaquin Valley College and the University of the Pacific
campuses in Stockton, California, the conference focused strongly on
interactive presentations with each member of the audience at a computer
(usually a Macintosh). Each contributing speaker was restricted to fifteen
minute discussions, and so the information presented about each topic was
general in nature and lacked depth. The invited lectures were longer (around
two hours) and so were much more useful and informative. Appendix I contains
the conference program.
The conference provided an opportunity to express both complaints about the
Maple package and some suggestions for improvement. I came away from the
conference with a new understanding of programming syntax, symbolic
manipulation functions, LaTEXreg. support, and plotting techniques.
Springer-Verlag displayed a collection of books related to Maple, several of
which I will be ordering.
Included in Appendix II, the list of book titles on display at the conference
contain a large number of Maple texts. In particular, Edwards' Advanced
Calculus is an excellent book on group theory techniques for solving
differential and integral equations, which I am ordering. Essential
Maple by Corless provides key information on using Maple effectively.
Describing common mistakes in syntax like the difference between " ` " and " `
", the significance of ":" in ":=", and the use of functions such as unapply,
the book is invaluable. Solving Problems in Scientific Computing Using
Maple and MATLAB by Gander is a collection of engineering problems with
their corresponding solutions in Maple and MATLAB code. By showing how to
solve a variety of problems with these programs, Gander has shown how one can
use Maple and MATLAB to their fullest.
The new version of Maple V, revision 4, is projected to be available before
the beginning of 1996. A beta version was available for conference attendees
to test on Windows 95 with a number of features disabled (animation and OLE
support). While revision 4 of Maple is not significantly different than
previous versions in its user interface on the UNIX and PC platforms, the
Macintosh version is supposed to be significantly improved due to numerous
complaints about its unfriendly interface. A piecewise function is new in the
version; it allows operations on piecewise smooth functions like integration
and differentiation removing the user's unwieldy dependence on Heaviside
functions for these types of functions. LSODE, or Livermore solver for
Systems of Ordinary Differential Equations, is to
be available on the new version. LSODE provides the capability for numerically
solving systems of stiff or pathological differential equations and is
currently superior to most other differential equation solvers. A number of
new graphics routines are being added to make it easier to plot results. For
example, three-dimensional plots will be able to be translated, rotated, and
combined together in a lattice on a page. The geometry package has been
completely rewritten to allow it to be interactive with other packages. The
linear algebra package, linalg, has new features for handling matrices.
Finally, a general operating speed increase is expected for all platforms due
to some revisions of the kernel.
A number of problems we've had with Maple were foremost on my mind when I
attended the conference. Chris Howlett, Maple vice-president for research and
development, Tom Lee, research project manager for Maple, and several
programmers were available to discuss problems and suggestions, and a
suggestion sheet was provided as well. I talked with them about the problems
we have been having and I was generally satisfied with their response. The
following list contains the problems and suggestions I spoke about and the
responses by Waterloo Software:
* Encapsulated PostScriptreg. support in the plots for Maple. I was informed
that the PostScript code generated by printing the plots is already compatible
with encapsulated PostScript, but the bounding box is probably too large for
any standard translation program to be able to handle. To fix this problem, a
programmer with Waterloo suggested the following fix: use the command
interface(plotoptions=`portrait, width=50, height=70`) before plotting to
resize the bounding box to a reasonable size.
* The PowerMac implementation of Maple V often crashes unexpectedly and
requires an unusually large amount of memory to function. I included this
complaint on the suggestion sheet, and one of the Waterloo programmers stated
that they were aware of the problem and were working on it for revision 4.
* Provide a root finder that doesn't skip roots in an interval (i.e., fix
`fsolve/setrange`). One of the programmers was intrigued with this problem.
He suggested I send in some information about the problem to support at
Waterloo.
* Stack-size problems with Jayakumar's plate equations. A programmer seemed
surprised that he was having problems. He didn't have an answer for me except
for suggesting to send the problem into support with information on the system
configuration.
* Three-dimensional graphics cannot be combined with two-dimensional plots.
The solution to this problem was discussed by Matthias Kawski at Arizona State
University who spoke about using op commands to pull vital information out of
the results of the two and three-dimensional plots and to combine them
together.
* Allow title and legends on plots to be edited easily. Supposed to be
included in the new version of Maple V (maybe).
* Eliminate maddening "worksheet" interface on Macintosh. To be eliminated in
revision 4.
* Revise dialog boxes in UNIX implementation of Maple. I didn't discuss this
topic with any of the programmers since it wasn't that important.
* Integration of functions over domains where the integrands are singular. I
included this on the suggestion sheet. I have since located a worksheet
located in the share directory that discusses this topic.
I have found a number of sources of information for help or discussion on
topics related to Maple. They are provided in the following list along with
other sources that I already have:
* www.maplesoft.com This web site is provided for information and
support from Waterloo software on Maple and related programs.
* www.eas.asu.edu:80/~asufc Information on developing
three-dimensional graphics with Maple is provided here.
* support@maplesoft.on.ca Maple's e-mail address for support.
* http://mathserv.math.sfu.ca/Software/MapleInternetList.html A list
of internet sites that discuss Maple-related issues.
Appendix I: Conference Program
Appendix II: Books on Display at Conference
Appendix III: Papers Delivered at Conference
Appendix IV: Public Relations Sheets on Maple V, Revision 3
Appendix V: Maple Newsletter