|
Professional Information
|
Background
I am a mathematician and computer
scientist. I have been working in parallel processing since the
1960's. If you would like to look at my resume, feel
free to browse there. You may also wish to view some of my published
papers.
Current Pursuits
2006
![[NEW!]](icon_gifs/new_animated.gif)
|
Stay tuned...
|
2005
|
This year started with a fairly frantic attempt to get the Division
Leadership Conference for the Grand Lodge of Virginia to take place at a
new location, the Taylor Middle School, in Warrenton, VA. This was a
much more central location in Division Six, making the average travel
distance from the 24 Lodges more than ten miles shorter, but some Lodges
that were near the previous locations in Stephens City and Middletown
had to drive considerably further. The facility didn't have the
auditorium available, so we used the cafeteria for the main session.
As long as the Most Worshipful Grand Master was happy...
This was also the last year for the Scottish Rite Degrees in the Pike
Recension. I discharged the part of the Venerable Master in the 20th
Degree for the last time at the Spring Reunion. I will miss doing that.
And I'm still the Grand Visitor and Lecturer of the Grand Council of
Cryptic Masons of DC, but I hear that there are those who would rather
have someone else doing the job--maybe someone who wouldn't be so
insistent on ritual improvement. Except that the degrees were done more
than once this year without my participation at all--couldn't be there
due to business travel.
Work: More travel (first time I was in Cincinnati in close to 25
years), more assignments, more deadlines. Lost a week of vacation due
to being at maximum accumulation (240 hours) for weeks during the late
spring and early summer. (Missed the Grand Council of Minnesota's last
meeting in St. Cloud and in the month of May, too.)
Music: Conducted my Xmas piece ("Christmas Overnight") at the December
concert. We'd played it fairly recently (2003), so I was surprised to
be doing it again. Maybe we need more music in the library? Well, I'm
still writing it...although there are about five pieces in the library
that Vienna Band has yet to perform.
More on music: Got into writing parody songs for the company. This
started when I wrote a parody of "When I Was a Lad" for the retirement
of the Center's vice-president, although that wasn't performed until
after I'd done a couple of songs about the pending office move (to the
fifth building in five years). Those were followed by songs about
spring cleaning (before the move), meetings that don't start on time,
the vicissitudes of Monday mornings, and so on.
|
2004
|
At the very end of 2003, I accepted yet another line position in a
Masonic body, that of Tiler of a Council of the Allied Masonic Degrees.
A commitment of advancing offices until 2010. "Committed" may be the
right word...
My activities as asst. conductor of the Vienna Community Band came to an
end with the notorious "Four Conductors--No Waiting" concert in March.
The three candidates for the position of Bandmaster each conducted four
pieces, while I conducted the opening ("Washington Grays") and closing
("Dawson Masonic March") numbers of the concert. Our new Bandmaster,
John St. Amour, assumed that position effective in September. I am
working with him on compositions and arrangements for performance by the
Band. At our fall concert, my setting of Lewandowski's "Mah Tovu" will
be on the program (under the title, "Balaam's Blessing").
Work has become somewhat more routine, as I improve my skills in the
assigned task.
|
Later in 2003
|
See the Verdict of History.
Job is still pretty much the same, but with more urgency. One
assignment required short-notice travel (Atlanta and Birmingham) in each
of three weeks this fall. I'm still trying to get caught up from
missing newspaper recycling for three weeks in a row! I did some
extracurricular work for another group relating to text processing,
using my expertise in that area and also in Visual Basic for
Applications. Still waiting to see how it works out; they're so busy,
they don't have time to save time with new techniques.
There is no end to my folly in accepting unpaid labor. I have now
agreed to serve as Assistant Conductor of the Vienna Community Band. My
debut will be at the Holiday Concert on 07 December 2003, conducting my
own piece, "Christmas Overnight." (Not a premiere; it was first done by
Vienna about six years back.)
More of the above. I also agreed to replace Right Worshipful Edmund
Cohen, now Grand Jr. Deacon of the Grand Lodge AF&AM of Virginia on
the Committee on Masonic Education and Publications, as Division Six
Provost. ("Provost" is the term used for upper level Masonic education
officers in Virginia.) No official visits, but Div. Six is far-flung,
unlike the urban one comprising Arlington, Alexandria, and inner Fairfax
County. I'm not sure how much travel I want to do for this...
|
2003 (and 2002)
|
No change in my employment. Same 'ol, same 'ol. Travel has mostly been
to Kansas City and Chicago, lately. Met some family in Chicago from a
remote branch of the family tree on my mother's mother's side; one of
them was doing genealogical research, which is how we got in touch. I
had some information that she didn't know or hadn't been able to
retrieve. She's traced a lot of the family back to Satoraljaujhely, a
town or small city in Hungary (then the Austro-Hungarian Empire) near
the border with Ukraine.
The George Washington Community Orchestra changed directors (replacing
Desi Alston) and stopped using people outside of students, I think.
They wanted to borrow my piccolo, but I insisted that I come with the
instrument or not at all. Too bad.
Flute
Talk magazine will be doing an article on community music groups in
the near future. One of the persons interviewed was yours truly. Maybe
even a photo will appear! Probably the May
2003 issue will contain this article.
Did my first theatre production in some
years: The God of Isaac, produced by the Temple Rodef Shalom
Players in the fall of 2002. I also came up with the logo design, mostly. (That one is my
design suggestion; I don't have a digital version of theirs, which
reverses the two masks. I think mine is slightly more amusing...)
Many developments in Freemasonry: Became a 33rd
in late 2001, named Cryptic Mason of the Year for 2001-02 in March of
2002, received the Grand Cross of Color from the International Order of
Rainbow for Girls in June of 2002, and was appointed General Grand
Musician of the General Grand Council of Cryptic Masons International in
October, 2002. Was installed as Worthy Patron of Hope Chapter
No. 73, OES in March, 2003. (It must be all downhill from here,
folks...)
Ever see an animated Tarot card?
|
2001
|
I'm still at MITRE.
I spent most of the year on two major assignments. Time schedules were
tight, which had its effects. But that's done. Time to move on.
The "events of 2001," as they have been called, cost the life of a good
Mason, who was also a sponsor employee with important responsibiltiies.
Alas, my Brother.
I rejoined the GWU Orchestra this spring, playing piccolo. And talked
the Vienna Band into designating me "Composer in Residence."
|
2000
|
I spent the early part of this year on proposal development...a bit
unhappily, as it turned out, because the customer really wanted to spend
the money on something entirely different from the officially-announced
topic of the procurement. Of course, that was still my fault, even
though I never even met the customer... Such is life in the government
contracting world.
So I changed jobs. I now work for The
MITRE Corporation in support of one of those government
organizations abbreviated to three letters. No, not the one you
think...
|
1999
|
In 1999, I engaged in a number of miscellaneous activities,
including pre-proposal abstracts, new-start briefings, and learning
about CYC. I also contributed to the
Campaign Assessment final report, including this graphic. I also built advisors for
purchasing scanners and digital cameras, implemented entirely in
JavaScript, and wrote a guide for design of web pages.
I've also been doing some art work for various papers we are producing.
Here is Li'l Robo. And this is a system administrator in the field. And this
is a space station. If you have a fast
connection, take a look at the UFO movie (3.3
MB). (The UFO animated GIF is 800KB but has
banding and is not as good looking.)
|
1998
|
In 1998, a new DARPA contract in
autonomous tactical systems began. We studied how to deal with robotics
and multiple robots in the presence of adversarial action. There are
many applicable technologies: AI planning, AI game playing, OR games
(discrete and differential), fuzzy methods, AI expert systems and
knowledge engineering, and so on. We weren't trying to address scene
recognition, sensor interpretation, actuators and effectors, and so on;
we hope that those problems will be solved by someone else. Lots of
interesting research topics! The first job is to assess the state of
the art in these technologies; then maybe build something. As a
proof-of-concept trial, one-on-one sailboat racing has been proposed for
late summer (but with a skipper advisor, not a real robot running the
tiller and trimming the sails).
|
1997 |
In 1997, I began working on DARPA
programs relating to systems architecture and security ("information
assurance") for command and control (C2) applications.
It's still a little early for results to arrive in what are called leading-edge services, which is a logical
(virtual?) testbed for migrating capabilities from R&D to deployment
as part of the Defense Information Infrastructure.
|
1996-7
|
In 1996-7, I worked in a number of miscellaneous areas of computer
science, primarily on government contracts. These included:
|
1996
|
Earlier in 1996, I did some proposal development work for a planned NIH program in application of parallel
processing to x-ray crystallography. Sorry, but the details are
proprietary.
|
1995
|
In 1995, I was working for Cray Research (not Cray Computer Corp., which
Seymour Cray founded after leaving Cray Research; Seymour died in an
automobile accident late in the summer of 1995). Primarily, I was
assisting in the marketing of a planned reduced-size vector computer
similar to the C90. A lot of people wanted to buy it, but no one wanted
to provide the capital to engineer it to the point of being
manufactured. So it didn't happen.
One of my responsibilities was to assist in an evaluation of its
capabilities on a problem in space-time adaptive processing (STAP) for
radar operating on a fast-moving platform (i.e., a jet fighter), which
is subject to phenomena not characteristic of fixed station radar or
radar on slow-moving platforms (ships). Some of the work centered
around the QR factorization of a symmetric matrix (another way of
looking at the Gram-Schmidt process), with which I was quite familiar.
In addition to writing a lot of the final report and planning the
presentation for the results (the program manager was new at this...), I
also developed a model, based on some of the work of my dissertation,
which was able to predict the performance and scalability of the core
computing kernel on a new machine, given its basic specifications. The
contract sponsor (MIT Lincoln Lab) found this rather impressive, as they
didn't know it was possible to create such analytic models. (Guess I
should have published more widely?) Our project results were
sufficiently remarkable that we won the competition for follow-on work.
But by then, SGI was getting ready to buy out Cray Research, many of us
were being thrown overboard (some reward for high-falutin' effort, eh?),
and I doubt that much success was had thereafter. Certainly, SGI didn't
manage to make Cray Resarch very successful as a subsidiary.
|
1994
|
The less said about 1994, the better... |
1992-3
|
During this period, I provided the technical leadership for a DARPA
contract on which an 8-A company provided facilities management and
consulting services for the transfer of parallel processing technology
to certain defense-related activities. At that time, the center,
sometimes called the Enterprise, was furnished with an Intel iWarp, an
Intel iPSC 860, and later a Connection Machine 5. For various complex
reasons, the software was not installed quite correctly, and Intel had
no record of the Enterprise as a customer, which made obtaining support
rather difficult for some months. Eventually, this was resolved, and
our job became easier. My principal activities included writing a user
guide for the Enterprise systems and building a demonstration program.
The demo used the iPSC 860 (and Sun front-end workstation) to compute
the magnitude of the earth's magnetic field, using a spherical harmonic
model developed at the National
Geophysical Data Center, using digital terrain elevation data; an
entire square degree of the earth's surface, consisting of 1.44 million
or 720,000 points (depending on latitude; i.e. latitude intervals of 5
seconds and longitude intervals of 5 seconds near the equator and 10
seconds near the poles) was computed in about three minutes or less. A
very large number of trig functions to be computed! With more recent
hardware, this would take much less time.
|
1989-92
|
During this period I worked as a system engineer on a defense-related
project. I had the opportunity to design some interesting programs (a
database with a graphical retrieval mechanism, e.g.) and do some
interesting analyses. I was even asked to design a complete follow-on
system at the technical engineering level in a field in which I had
virtually no experience. Thanks to a good tutorial someone found for
me, I was able to put together the complete specifications in about six
weeks. I also ran a statistical analysis of a competition among vendors
for a particular software system, which revealed that one of the
programs performed no better than random on the test problems. (I think
that was the vendor the government wanted to win, so that may have been
a bit embarrassing!) Any more details about this work are proprietary,
however.
|
Recent Completed Work
My most recently completed work in high-performance computing consisted
in parallelizing an application in geophysics: The computation of the
geomagnetic field according to a spherical harmonic model of field
components. The original model was created at the National Geophysical Data Center in
Boulder, CO. The parallel version uses that basic model in combination
with digital terrain elevation data (DTED) to compute the field strength
over an entire square degree of the earth's surface at intervals as fine
as 3 seconds of arc, or up to 1.44 million points. On an Intel iPSC/860 with 32 processors, this
task required approximately three minutes and included computation of
false color imagery to represent the intensity. This work was supported
by DARPA under the HPCC Initiative.
Other activities in the mid-1990s included a number of activities
related to signal processing, particularly a study funded by MIT Lincoln Laboratories on
space-time adaptive processing (STAP). Preliminary results from this
effort indicate that this technique, long known but computationally
infeasible on the computing equipment of the 1970s, is now likely to be
a viable approach for radar systems of the near future.
In 1991, I published, jointly with Eric Opp and Mark Cullen, a paper in DMCC 6 (Sixth Distributed Memory
and Concurrent Computing Conference, held in Portland, OR) demonstrating
that high-performance adaptive optics, at a rate of adjustment of about
1 kHz, was feasible using the technology available at that time
(actually, the work was done in 1989). Most articles on the subject
seem to claim that this level of performance wasn't achievable until
considerably later and wasn't declassified at all until the very late
1990s or Y2K. Our paper was unclassified, as was most of the work on
the ground-based laser of the Strategic Defense Initiative. Don't
believe everything you read in Tom Clancy! (I.e., The Cardinal of
the Kremlin.)
If you can't afford a supercomputer but have a lot of workstations
around, you might want to explore PVM: Parallel Virtual
Machine, which is a tool for developing parallel/distributed
programs on networks of workstations, as well as on various
supercomputing architectures.
Colleges and Universities
If you read my resume, you'll notice that I went to Brown University and some of the alumni home pages may
be of interest. Brown University alumni/ae may wish to help Save the Podes!!! I also spent time at the
Twin Cities Campus of the University of
Minnesota (my elementary school and high school were there, and I
took about half a year of college courses in calculus, German lit., and
social science before going to Brown), at New York University--Courant Institute of
Mathematical Sciences (CIMS), where I received my doctorate, and at the
College (now University) of
St. Thomas in St. Paul, Minnesota, where I received my
MBA.
Economics and Politics
In an area unrelated to high-performance computing, I am also a Visiting
Scholar of and regular participant in the Colloquia on Classical Liberal
Thought at the Locke
Institute at George Mason
University in Fairfax, VA. A recent speaker at a Colloquium was
William Niskanen, CEO of the Cato
Institute.
I served as precinct captain for Blake precinct in the fall 1997
campaign for a Virginia House of Delegates seat by Jeannemarie
Devolites. The expected margin of loss in that precinct was about
150-200 votes; the actual margin was only 37. About thirty votes of
that improved margin were probably due to our efforts in the precinct;
the rest were due to an apathetic job by the opposition who failed to
get out the vote in a secure precinct. Jeannemarie won the election by
several percent of the vote, turning out an incumbent.
|
 Reference
|
- HTML
Primer
- Introduction
to HTML
- HTML FAQ
- HTML
Quick Reference
-
Netscape HTML Extensions
- HTML
Forms Tutorial
-
HTML+
- HTML
3.0
- HTML 3.2
Fact Sheet
- HTML 3.2
Reference Specification
- Weblint (an HTML syntax
checker)
- Emacs html editing mode (my version,
with additions to the original created at NCSA; this is the actual code,
which you may download and use--byte-compiling is advised! Be sure to
remove the ".txt" suffix.)
- Adds horizontal rule (HR) and line break (BR) tags
- Updated 02 Aug Wed Nov 14 10:00:50 EST 2001
entering an IMG (to keep lynx users and those that don't download every
image happy)
- 03 Feb 97 update adds FONT tag (but you have to add your own SIZE
and COLOR attributes)
- 03 Feb 97 update adds table mode
- 21 Mar 97 update adds ability to add any container tag to region,
extends table mode, and changes IMG source entry to assume a local file
(rather than just a string entry--allows completion, but editing must be
done on the result; is this a good idea?), slightly modifies table mode
(add caption becomes C-c C-t k, so that C-c C-t c adds column spec and
C-c C-t g adds a column group)
- 7 Apr 97 update adds C-c SPC command to insert non-breaking space
special character ( ) which is handy in tables
- 19 Sep 97 update fixes bugs with html-add-superscript and
html-add-subscript (they opened an area rather than a field; the latter
is correct) and generates a level-1 heading for the title of the
document when html-init is used (as well as putting the title into the
<HEAD> area)
- 22 Sep 97 update removes deprecated XMP and LISTING functions
(since the same capability is provided by the <PRE> tag)
- 25 Nov 97 update adds !DOCTYPE insertion to html-init function
- 03 Dec 97 update adds entry of JavaScript areas with C-c j command
- 11 May 99 update changes various fixed-width commands (for the
<CODE>, <KBD>, <SAMP>, and <TT> commands) to be
subcommands of C-c w.
- 11 May 99 update adds html-new-paragraph-mode variable to control
whether C-c p adds <P> only or adds <P> and </P> and
then leaves point between them. Default (nil) is the old style.
- 11 May 99 update changes C-c SPC to allow a prefix argument for
adding multiple non-breaking spaces with a single command.
- 13 May 99 update adds form commands under C-c C-f:
| keystroke |
command |
function |
| C-c C-f f |
html-add-form |
Add a FORM area |
| C-c C-f a |
html-add-form-textarea |
Add an INPUT TYPE=TEXTAREA |
| C-c C-f b |
html-add-form-button |
Add an INPUT TYPE=BUTTON (for use with JavaScript) |
| C-c C-f c |
html-add-form-checkbox |
Add an INPUT TYPE=CHECKBOX |
| C-c C-f e |
html-add-form-reset |
Add an INPUT TYPE=RESET (reset button) |
| C-c C-f h |
html-add-form-hidden |
Add an INPUT TYPE=HIDDEN |
| C-c C-f o |
html-add-form-option |
Add an OPTION (to a SELECT area) |
| C-c C-f p |
html-add-form-password |
Add an INPUT TYPE=PASSWORD |
| C-c C-f r |
html-add-form-radio |
Add an INPUT TYPE=RADIO (radio button) |
| C-c C-f s |
html-add-form-select |
Add a SELECT area (values supplied by OPTION specifications) |
| C-c C-f t |
html-add-form-text |
Add an INPUT TYPE=TEXT |
| C-c C-f x |
html-add-form-submit |
Add an INPUT TYPE=SUBMIT |
- 21 July 99 update enhances C-c f (html-add-font) to allow user to
enter size/color/face specifications.
- 12 Oct 99 update adds underline (C-c C-u) and strikethrough (C-c
C-x) text capabilities; allows html-add-line-break (C-c k) to have a
prefix arg for entering multiple line breaks.
- 21 Nov 01 update adds optional parameter entry for table row
(valign, halign) and table datum (width) if there is a prefix argument
(i.e., C-u before). Completion is provided for text values.
- 29 Nov 01 update to move towards XHTML: lower-case tags, matching
closing tags or closing mark, etc. Functions html-add-head,
html-add-body, and html-add-html are no longer interactive (i.e., not
available with M-x); one should use html-init to set up a new HTML file.
For new files, the default setting of html-new-paragraph-mode will be T,
rather than NIL; this will be enforced by a "Local variables" section
added at the end of the file to make sure it is set when the file is
opened for editing (an XHTML requirement).
- 07 Dec 01 update to add function html-add-meta.
An emacs Web browser
is available for experimentation. Follow that link if you want to find
links for where to download an emacs editor for your system (even the
Mac), learn about emacs-lisp, or explore other GNU software links.
If you have mastered the above, you might want to look at Advanced Web
Development information written up at Georgia Tech. (The site seems
to have moved to a new
location, however, and may no longer be updated as developments
occur.) The Web Reference
pages are also very useful.
For Mac users, the clip2gif
home page provides access to a really clever (clipboard sensitive,
scriptable, etc.) way to create GIFs for your illustrations. More
information can be found in the Transparent/Interlaced
GIF page (if it is working). Also look for a program by M. Piguet
called Gif Builder if you would like to do some simple Web page
animation.
Another place for information on Web animation is provided by CNET.
Mac users (like me) might also like to visit some of the following
sites, derived from a recent issue of MacWEEK:
For those who are into programming for the Web, the NBS Unix World Wide
Web Management Utilities provide some useful tools and techniques.
In addition to CGI programming, one may also look into server-side
includes (SSIs) as ways to accomplish some tasks in a simpler manner.
Beyond HTML, there is JAVA. Some
Java
scripts can be borrowed and adapted. CNET also has a Java
resource page.
CNET
offers some thoughts on practical uses for JAVA.
|
![[color photo of me]](roger01d-148x240.jpg)
![[welcome door]](door.gif)
![[SoundFileIcon]](icon_gifs/soundfile_animated.gif){kind=icon}
Welcome! (41K sound) This home page is 
.
(Aren't they all?)
![[mailbox]](icon_gifs/anim_mbox.gif){kind=icon}
28
Dec 1994
Thu Mar 23 08:55:29 EST 2006
![[new!]](icon_gifs/new_tiny.gif){kind=icon}
![[dead link]](icon_gifs/RIPicon.gif){kind=icon}
