I'm a computer scientist,
which means I usually design and implement computer software.
Since I've had faculty appointments at the
University of Puget Sound
and Colorado State University,
it also means I've taught others about computer science.
I worked at lots of interesting places and
on the side I write software and text for the Linux/Open Source community.
You can read my resume for details.
All this has earned me some grey hairs--as seen
in this recent photo.
Although I'm most interested in operating system work (drivers, networking, system administration, etc.), I also have written microcode, large database conversion programs, and performed system management under contract. You could say I work from the very small to the very large scale problems.
I've worked with C for a long time since it's effective and
nearly universally available.
C++ is an "interesting" extension to C,
but I've seen how it's possible to create
a functional but really opaque library written in C++.
That may be an argument for Java, except for performance issues.
My current language of choice is Python. I've long been
intrigued by more pure functional programming languages.
My dissertation examined how John Backus'
FP language could be implemented on a shared memory multi-processor.
I've been casually exploring how I can use Haskell,
and I've also been looking into Fortress.
The script du.py determines the sizes of individual files and the accumulated storage attributed to files below each sub-directory (in kibibytes). It then sorts these lists and formats them to show the greater users of the disk first while displaying the hierarchy.
I've released SANE drivers for two models of Fujitsu scanners (M3096G and ScanPartner 15C). You can fetch them from .... I welcome your comments or enhancements regarding these programs.
I've long been interested in satellite based navigation systems. My first exposure was when I worked weekends with off-duty sailors who operated one of the ground stations for the Transit system. A few years later I worked with a fellow who told me of his experience programming a DEC PDP-11/Transit based navigation system used on commercial freighters. Shortly after GPS navigation came into existance I bought a Garmin receiver. I continue to make modest improvements to an upload/download program for a number of Garmin handheld GPS receivers. It can be found at here. I've also written programs to plot the tracks on geo-referenced NOAA/USGS charts. (This code will be available in a bit, but drop me a note if you can't wait.)
Besides writing stuff for computers to execute, I've written stuff for people to read. Most of my recent published work has been directed towards the Linux community. My report on a survey I did of Linux kernel developers appears in the Linux Gazette and was reprised in the Linux Journal. My book, Inside Linux: a Look at Operating System Development, has sold out, but I'm told it's available on the web.
The Ricochet wireless modem was the forerunner of the current 3G and 4G wireless data network. I've written an article for the January 1998 issue of the Linux Journal which describes my experiences with this device. I've explored their coverage and have partial maps of their pole-top repeaters in Seattle region and the San Francisco Bay area. (The blue dots are Ricochet's Wire Access Points, the red and green dots are the pole tops I've identified, and the light-blue/cyan dots are some misc points.) It was a good idea, but Metricom has folded and I'm interested in seeing if anyone picks up the pieces.
I wrote an outline of the early history of Linux development which appeared in the March 1995 issue of Linux Journal (as well as the Linux Sampler). It also describes the dual development path for production and experimental versions of the kernel.
An earlier article (September 1994) described my effort to write a device driver for the Cyclades Cyclom-8Y serial asynchronous multiplexer.
Besides articles, I've written a book review and conducted an interview for the Linux Journal.
My writing activities have brought me in contact with the National Writers Union.
The network time protocol provides
a mechanism to synchronize computer clocks to national standards.
The NTP adjusts how the computer's internal elements are used to
advance the time-of-day clock. These adjustments are recorded and I've
found that when I examine how the adjustments change over time, they
display the effect the room's temperature has on the computer's crystal
clock!
I installed NTP on my systems as soon as a could. Later I installed
the TAPR (Tucson Amateur Packet Radio) TAC (Totally Accurate Clock)
which interfaced a GPS receiver to a NTP enabled computer. I had this
running for a while (until the receiver's cover failed and rain got to
the electronics). More recently I've deployed Garmin 17 and Garmin 18PC
receivers which can directly interface with my computers. (I've found
that I can get a good-enough signal through the roof so I don't have
to worry about the rain. :-) With these sources in place, two of my
systems qualify as NTP stratum 1 servers for the rest of my systems.
My systems visible to the Internet (including this server)
thus are NTP stratum 2 servers.
I gave a talk about my experiences with NTP to the Greater Seattle
Linux User Group. It's available as a
video
and you can view my presentation
pages
I've been involved in the local Unix users group and the local Linux users group since its inception.
I've been sailing for decades. It was only natural that I moved to the Puget Sound area -- it's got a temperate climate, attractive cities, and great cruising. We do our cruising on S/V Cheshire Cat.
Sometimes these pasttimes come together. The Corinthian Yacht Club has a page on the web.
My compulsive reading has led me to visit a great many bookstores. This is but a partial listing I developed in the mid-1990s. I'd appreciate any additions folks can offer. (There's no way I'll ever complete a catalog of my books.)
I've developed an appreciation of classical music over the past twenty years. My tastes run from Baroque to Romantic. Some day I may get my CD collection on-line. The chances are better than for my book collection.
I've attended a few meetings of the "webmasters" group of the Seattle Community Network. Check out their page.
My full name is Randolph Allen Bentson, but for the sake of search engines, it's worth noting I'm often called Randy Bentson. Bengt Nilsson was a distant ancestor, from which the family name Bengtsson was derived. Upon immigration to the U.S., this name was shortened to Bentson (and Benson by some relatives). The given name, Bengt, is the Swedish equivalent to Benedict, which could be interpreted as "blessed". If you're a descendent of Bengt Nilsson and Ellna Sonesdotter, I'd welcome the contact.
Feel free to drop me a note at
bentson _AT_ holmsjoen _DOT_ com
