NA Digest Saturday, February 22, 1992 Volume 92 : Issue 8

Today's Editor:

Cleve Moler
The MathWorks, Inc.

Submissions for NA Digest:

Mail to

Information about NA-NET:

Mail to


From: Jack Dongarra <>
Date: Sat, 22 Feb 92 09:28:49 -0500
Subject: New Version of NETLIB at

We just installed a new version of netlib at
This version allows a user to set the chunk size for retrieving
large files, so things don't get split up into pieces. As an example,
to receive the postscript file of the Linpack Benchmark Report
in one mail message you could send the following two lines in a mail
message to

mailsize 450k
send performance from benchmark

you would then receive one file with the information.
(Unfortunately not all gateways allow large messages. If there is a
gateway between and your machine that doesn't allow
messages of this length, you will have to use the default mode of
message splitting. However, the recombining has been made simpler:
a user needs to strip off the mail headers and ``sh'' the resulting
shar files, in any order, to recompose the original file.)

There are a few additional new features, try: ``send index''
for all the details.
Jack and Bill


From: Jeff Scroggs <>
Date: Mon, 17 Feb 92 15:57:04 -0500
Subject: Information on CFD Codes Sought


I would be pleased to receive information and opinions on the currently
available public domain CFD codes. The fluid flow problem includes
the following attributes (the complete description is too long to include):

1. Flow is subsonic (slow).
2. There are regions of compressible flow caused by large
temperature gradients.
3. Soret (multi-component) diffusion should already be modeled,
or be easily added.
4. Irregular geometries are a must, but we have grid generation
routines available.

The code should be well documented and fairly robust (graphics
not necessary).

Your cooperation would be appreciated,
Jeff Scroggs

US Mail: Department of Mathematics
Box 8205
North Carolina State University
Raleigh, NC 27695
AT&T: (919)515-7817
FAX: (919)515-3798


From: Warren Ferguson <>
Date: Tue, 18 Feb 92 17:48:18 CST
Subject: TEX Torture Test

Does anyone know where I can get a torture test for TEX. I know Knuth
mentions such a text for TEX 82, but I've never seen the test. Can
anyone provide the test to me, or tell me where I can get the test?

Warren Ferguson


From: Art Werschulz <>
Date: Thu, 20 Feb 92 09:53:25 EST
Subject: CAC-NET (Continuous Algorithms and Complexity)


CAC-NET is an electronic conferencing system for people interested in
continuous algorithms and complexity. It is similar to the NA-NET
system that has been successfully used by the numerical analysis community.

Continuous algorithms and complexity is a rapidly growing field which
studies algorithms and computational complexity for continuous
mathematical problems which often arise in science and engineering.
By computational complexity we mean the intrinsic difficulty and
optimal algorithms for solving such problems. Examples of the
problems that are being studied include: ordinary and partial
differential equations, continuous optimization, multivariate
integration and approximation, matrix multiplication, and systems of
polynomial equations. Applications of continuous algorithms and
complexity to various disciplines are also of interest.

Material appropriate for distribution on CAC-NET include:

Announcement of research results

Discussion of research questions

Report and paper availability

Announcement of forthcoming Journal articles

Address changes

Announcement of available positions

Forthcoming meetings

Material is distributed to the CAC-NET in the form of a newsletter,
the CAC Digest. You may submit material to the CAC Digest in the form
of an electronic mail message addressed to

To receive information on CAC-NET, including information on how to join,
send mail to This is an Internet address.
If you are on Bitnet, please make the appropriate adjustment. If you
have questions or comments send mail to
Please note that, depending on network delays, you may not receive a
reply immediately. If you are unable to reach CAC-NET you can send
mail to
Art Werschulz
450 Computer Science Building
Columbia University
New York, NY 10027
for additional information.

Art Werschulz
ATTnet: Columbia University (212) 854-7119 854-2736
Fordham University (212) 636-6325


From: John Reid <>
Date: Fri, 21 Feb 92 11:52:05 GMT
Subject: The Responsibilities of Editors to Referees

There is another matter with respect to refereeing that I would like to
raise. It concerns the responsibilities of editors towards their
referees. I think that if there is a disagreement between referees or
if an editor disagrees with a report, the editor should attempt to
achieve consensus by sending referees copies of other reports. Of
course referees are sometimes wrong (even me), but majority voting is
not the way to procede - one referee may have found a snag that the
others have overlooked. Of course, in the last resort the editor may
have to reach a decision between conflicting advice, but the referee
whose advice has not been taken should feel that he or she has had
every opportunity to present his or her view and should understand that
there is scope for a difference of opinion.

I have to admit that my reason for raising this is personal. An editor
recently decided to ignore my advice. It is clear from subsequent
correspondence that he did not even understand the point that I was
making. Probably I did not explain it adequately, but he did not even
try to reach consensus. I have now returned another paper that he has
sent to me. Why should I spend time writing a report that is likely to
be ignored?

I am an associate editor myself of two journals. With this hat on, I
feel that my referees need to be cosseted. Many years ago, I was sent
a copy of the other referee's report, and found this interesting. Since
then I have always copied my corresponence with the author to the
referees, including copies of other referees reports.

John Reid,
Atlas Centre,
Rutherford Appleton Laboratory,
Oxon OX11 0QX


From: Robertson J.S. Maj <>
Date: Mon, 17 Feb 92 13:25:48 EST
Subject: Fortran Tools

This is a brief followup to the note in Digest vol. 92, #7.

I'm the keeper of the Fortran Tools mailing list. It is rather
dormant, but what the heck.

Please drop me a line if you'd like to be added. To post to this
unmoderated list, send to

Also, there are several ftp sites which carry the tools:


Location: /pub/usenet/comp.archives/languages/fortran
DIRECTORY drwxrwxr-x 384 Nov 29 10:14 toolpack


Location: /programming
DIRECTORY drwxr-xr-x 512 Jan 9 1990 toolpack


Location: /pub/comp.archives
DIRECTORY drwxrwxrwx 512 Mar 8 1991 toolpack


Location: /
DIRECTORY drwxr-xr-x 1024 Apr 13 1991 toolpack


Location: /pub
DIRECTORY dr-xr-xr-x 1024 Oct 19 22:58 toolpack


Location: /.a/sranha-bp/arch/arch/comp.archives/languages/fortran
DIRECTORY drwxr-xr-x 512 Feb 20 1991 toolpack


Location: /pub/utils
DIRECTORY drwxrwxr-x 512 May 14 1991 toolpack

This list was obtained from the archie server at McGill University and
is accurate as of 17 Feb 92.

For those interested in porting the tools to the RS/6000 platform,
I've hacked the standard makefile to work with the xlf compiler.
Contact me directly for a copy.

Major J.S. Robertson
Department of Mathematical Sciences (914) 938-2453
USMA, West Point, NY 10996-1786 FAX (914) 938-2409


From: Gene Golub <>
Date: Sun, 16 Feb 92 13:24:34 CST
Subject: Parallel Circus XI: Preliminary Announcement


XI Parallel Circus
Location: Minneapolis, Minnesota
Dates: April 24-25

The next Parallel Circus will be held at the University of Minnesota
and will be sponsored by the Computer Science Department, Cray
Research Inc, and Minnesota Supercomputer Institute. The dates for the
meeting are Friday, Saturday April 24, 25.

What is the Parallel Circus? It is an informal meeting (the agenda is
organized at the meeting) of persons engaged in scientific computing
with emphasis on parallel and vector computing.

The circus is unique in that it is VERY informal, and thus allows us
to talk about the very latest results as well as interesting work in
progress. In our previous meetings there has been lots of informal
discussion and a very healthy mix of industrial and academic
participants. We expect a very exciting meeting since there is such an
extensive group of researchers in the Minneapolis area.

Details concerning the place of meeting and accomodations will soon be

There is modest support from the National Science Foundation for student
travel to the Parallel Circus XI. Those students requesting support should
send a one page proposal to Circus 92 ( giving the
reasons for attending the meeting, and a budget for expenses. The
student(s) should indicate their research interests and plans. A
letter verifying that the student is in good standing should be sent
independently by a faculty adviser. This letter should give the
student's GPA. We will be pleased to consider joint proposals which
would include the expenses of several students.

The deadline is 7pm (CST) on March 23 and we will try to get a
response by March 28. Correspondence by e-mail is desirable.

Organizers: Gene Golub, Bill Harrad, Ahmed Sameh, Apostolos Gerasoulis
Local Committee: Don Truhlar, Michael Olesen

FAX: (612) 625-0572 (Write "Student Support" on Fax.)


From: Adam Bojanczyk <>
Date: Sun, 16 Feb 92 15:03:47 CST
Subject: IMA Workshop on Linear Algebra in Signal Processing

April 6 - 10
Minneapolis, Minnesota

The Institute for Mathematics and Its Applications at
the University of Minnesota will sponsor a workshop on
linear algebra in signal processing. There are five themes
planned for this workshop with several invited and
contributed talks addressing the themes. The invited
talks are grouped according to the workshop themes:

(G. Golub, F. Luk, D. Tufts)

(R. Coifman, G. Strang, M. Vetterli)

(S. Haykin, A. Steinhardt)

(T. Kailath, I. Koltracht, P. Comon)

(S.Y. Kung, R. Schreiber)

There will be time set aside for informal discussions and exchange
of ideas between signal processing practitioners and numerical
linear algebra specialists.

Questions about the workshop can be directed to Adam Bojanczyk
( or George Cybenko (
For details about local arrangements please contact Willard
Miller, IMA, University of Minnesota, Minneapolis, MN 55455,
(e-mail or phone (612) 624 6066).


From: Ivan Dimov <"bgcict::ivdimov">
Date: Mon, 17 Feb 1992 13:09:02 EST
Subject: Workshop in Bulgaria


The second WORKSHOP on PARALLEL ALGORITHMS 92 (WPA'92) will be
organized at the Center of Informatics and Computer Technology
at Sofia, Bulgaria from August 18, 1992 to August 22, 1992.

The Workshop program will cover the following areas:

Monte Carlo Methods
Numerical Parallel Algorithms
Combinatorial Parallel Algorithms
Solving Mathematical Problems on Parallel Computers
Systolic Arrays

The address of Organizing Committee is:

Dr. Ivan Dimov -- WPA'92
Center of Informatics & Computer Technology
Bulgarian Academy of Sciences
Acad. G. Bontchev Str. 25-A
1113 Sofia, Bulgaria
phone: +359 2 708494;
telex:22056 KZIIT BG;
fax: +359 2 707273.

Authors are invited to submit 2 copies of an extended abstract.
Accepted papers will be published in the Proceedings of WPA'92.

Deadline for submission: 20 of April, 1992,
Notification of acceptance: 30 of May, 1992,
Final version due: 22 of August, 1992.

In case you are interested in more details and you are able to take part in
WPA'92 please contact Organizing Commettee.


From: Stephen Wright <>
Date: Wed, 19 Feb 92 21:23:24 CST
Subject: Postdoctoral Position at Argonne National Laboratory

Postdoctoral Position
Mathematics and Computer Science Division
Argonne National Laboratory

We invite applications for a postdoctoral research position
at Argonne National Laboratory. Candidates must have experience
in developing and implementing parallel numerical algorithms,
as demonstrated through research accomplishments in numerical
analysis or scientific computing. The appointment is for a one-year
term (renewable) and is available immediately.

The successful candidate will participate in a high-performance
computing research project whose aim is to develop and apply
algorithms and software for the solution of high-temperature
superconductivity problems on massively parallel computers.
The individual will interact with applied mathematicians,
computer scientists, and materials scientists. Project members
have access to state-of-the-art computing facilities, including
a 528-processor Intel Touchstone DELTA.

Familiarity with numerical modeling of phenomena in physics or
applications science and experience in obtaining computational
solutions to such problems are desirable.

Applications must be addressed to

Nancy Griparis, Box mcs-postdoc, Employment and Placement,
Argonne National Laboratory, 9700 S. Cass Avenue, Argonne, IL 60439,

and must include a resume' and the names and addresses of three
references. For further information, contact

Steve Wright (708-252-7847;

Argonne is an affirmative action/equal opportunity employer.


From: SIAM <>
Date: Wed, 19 Feb 92 13:24:01 EST
Subject: Contents: SIAM Numerical Analysis

Tentative Contents
SIAM J. Numer. Anal.
June 1992, Vol. 29, No. 3

Convergence of a Finite Difference Scheme for the Navier-Stokes Equations Using
Vorticity Boundary Conditions
Thomas Y. Hou and Brian T.R. Wetton

Method of Lines for Hyperbolic Differential Equations
Heinz-Otto Kreiss and Godela Scherer

A Multigrid Algorithm for the Lowest Order Raviart-Thomas Mixed Triangular
Finite Element Method
Susanne C. Brenner

A Spectral Collocation Technique for the Solution of the Wigner-Poisson Problem
Christian Ringhofer

Analytical Linear Numerical Stability Conditions for an Anisotropic Three-
Dimensional Advection-Diffusion Equation
J.M. Beckers

Analysis of a Finite Element Method for Maxwell's Equations
Peter Monk

Local Error Estimates for a Finite Element Method for Hyperbolic and Convection-
Diffusion Equations
Richard S. Falk and Gerard R. Richter

Orthogonal Spline Collocation Methods for Some Partial Integro-Differential
Yi Yan and Graeme Fairweather

Finite Element Error Estimates for a Blast Furnace Gas Flow Problem
S.-S. Chow

Numerical Integration of the Differntial Riccati Equation and Some Related
Luca Dieci

Approximation by Solutions of the Planar Wave Equation
M.V. Golitschek and W.A. Light

The Method of Resultants for Computing Real Solutions of Polynomial Systems
Eugene L. Allgower, Kurt Georg, and Rick Miranda

Image Selective Smoothing and Edge Detection by Nonlinear Diffusion. II
Luis Alvarez, Pierre-Louis Lions, and Jean-Michel Morel

A Viscosity Solutions Approach to Shape-From-Shading
Elisabeth Ruoy and Agnes Tourin

Numerical Methods for Inverse Singular Value Problems
Moody T. Chu


End of NA Digest