The LINCS/NLTSS Chronicles
James E. (Jed) Donnelley
These cartoons were drawn largely by
Bill Aimonetti (signature "BA") during a period of work on the LINCS/NLTSS
system when the system was in transition to full production use. During this
time, weekly meetings were held to talk about progress, problems, etc. Many of
the cartoons are dated. The dates range from August 1986 to November
1987. After the meetings, Bill often drew up a cartoon that portrayed some
highlight of the meeting. For each cartoon, I've created a thumbnail image and
I've included some text describing what I remember about the cartoon's content,
augmented with comments from others, especially Bill Aimonetti himself (thanks
Bill!). The thumbnail is a link to a higher resolution version of the cartoon.
I've somewhat sanitized the cartoons from their earlier form. Because of this,
there are some blank spaces in some of the cartoons. The cartoons are presented
in an order that I chose for no particular reason. In particular, there is no
attempt to list them chronologically.
This Web content is a working draft. If you have any familiarity with this
time period during the LINCS/NLTSS devlopment and you have any thoughts you'd
like to share for inclusion, perhaps you recognize some of the people in the
cartoons, maybe you see some errors in the text, have some criticisms, whatever--
please let us know at the address below.
This first cartoon is my favorite. It gives an idea how crazy
things got at some points during the development. In this cartoon, you can see a
code system on the left (the car) approaching the NLTSS system under
development. When Bill Aimonetti first started working on scientific code
systems, the work was described by Jan Moura and by Mike Pratt as "changing the
tire on a moving car." Add the additional complication of an operating system
under development, and this cartoon becomes a natural. You can see both the code
system and NLTSS are being worked on. In the car, standing up, is Jan Moura
leading the programming on the code system. The guys working on the code system
could be any of Bill Aimonetti, Dale Nielsen, or Scott Brandon. Of course, one of
them could also be driving. I'll leave that to your imagination. Working on
NLTSS are Jed Donnelley (in the hat) and Dave Fisher--who was working on the
Process server at the time. The signature is BA 8/86.
This cartoon depicts a party celebrating the the User Systems
Division (USD) and the Network Systems Division (NSD) successfully getting NLTSS
into production, running the scientific workload of the Livermore Computer Center
(LCC). That's a code system peeking out from behind the trees. The hat behind
the trees might be Harry Nelson or perhaps Pat Gray as the party was held at
Pat's house and used his pool. I (Jed) am the one running in the hat off the
diving board. Bill Aimonetti is the one trying to pull himself up on the board.
Bill and I both worked on that trick during the party. Mike Pratt is at the
lower right with the cork landing on his head. Bob Cooper us at the lower left
opening the bottle of champagne. I'm not sure who the rest are so there's some
room for imagination there.
This is a fun cartoon that depicts some of the tensions between
production and system development fairly late into the deployment of NLTSS, when
many production systems were running on it. I think that's Pat Gray on the right
trying to pull "the Clock" over to 10, probably 10 days without a crash. I think
that's Bob Cooper, on the left, with his fingers crossed and biting his finger
nails hoping that the various production codes will run to completion. You can
see the various production systems pictured. I've forgotten about the "PAD"
sheets and the name of the guy carrying them. At the lower left, Barbara is
trying to hold me (Jed) back from putting in some systems changes that might
destabilize the system. The signature is BA 8/24/87.
This cartoon shows Chris Hendrickson's bad dream as a bunch of the NLTSS folks head
off to Rockwell International to bring up an instance of the NLTSS system there.
I (Jed) am driving the chariot. I'm not sure who that is to my right. Donna
Mecozzi? Barbara Atkinson? Also in the Chariot are Jim Minton (behind me) and
Harry Nelson peeking out near the bottom. Pat Gray is waving good-bye. The
signature seems to be just BA, no date.
This is an example if a cartoon depicting one of the many
problems we ran into during the deployment of NLTSS/LINCS. This particular
problem was called "Bits from outer space" (at the time) because bits seemed to
be set to 1s randomly when they shouldn't be. Linnea Cook, who discovered the
problem, is at the keyboard and I believe that's Harry Nelson as the Darth Vader
figure behind her. The signature is MP/BA - no date. MP?
As I recall the "Bits from outer space" problem soon came to be known as the "12th
Bit problem" as it was always the 12bit in a word that was set to 1. Eventually
the problem was isolated to a hardware problem in a disk controller that was
fixed by Cray. The cartoon shows Dave Fisher (at the left), me (Jed - I think
with the hat and mustache), Linnea Cook (happy that "her" problem has been
solved), then perhaps Jim Minton and I'm guessing the last may be Rob Haynes
from USG. The signature is MP/BA87.
This cartoon depicts an example problem transiting from development to production.
I'm sure Jim Minton will love seeing this one and remembering it. Early during
the development of NLTSS, we found it convenient to be able to simply clear the
emulated disks in the "Encapsulated" system (sort of a virtual machine). Jim
Minton, who developed many of our support tools, set up a mechanism where just
by hitting the "F7" key the disks would be cleared. This function was passed on,
from system to system, even after we started running with real disks and into
production. In this picture, we see Jim having a bad dream after he had
inadvertently pushed that F7 key and wiped clean a bunch of DD29 disks on a
running system. In the background, you see Barbara Atkinson (tap, tap) and me
(Jed) with the tar and feathers. The signature (on the plunger) appears to be BA 5.
One of the big milestones in the NLTSS development was "Big P" production. I've
forgotten at this point just what the requirement was. Some up-time requirement,
no doubt. In this cartoon, you can see many of us chasing the Big P along the
calendar during late May 1987. I think I recognize all the characters in this
one. Leading the charge is Barbara Atkinson. In the next row are, left to right,
Pat Gray, Dave Fisher, and me (Jed). In the next row, again left to right, are
Gail Whitten (who worked on the "QAlib" test suite--hence the cannon), Bob
Cooper, and Norm Samuelson. Following is Chris Hendrickson with Harry Nelson
(and the snail) trailing the pack. Thats a code system flying overhead. The
signature is BA 5/17/87.
This cartoon is about another problem that plagued us for a
time. I really don't know who that is pointing the pistol. I see "TIM" written
down the holster and "AX" on the belt. Perhaps that was Tim Axelrod? I remember
the name, but I don't remember him having anything to do with this problem. From
the cartoon, it appears the kernel of the system was implicated. The signature is
Of course, after we deal with a problem we get to have a
celebration. In this case, pictures as mourning over the grave of "Mysterious
Death" while our shadows celebrate in the background. I believe in the picture,
left to right, are me (Jed), Chris Hendrickson (though perhaps that's Rob
Haynes?), Harry Nelson, Dave Fisher, and Jim Minton. The signature (to the right
of the gravestone) is BA 86.
Of course, while putting a system like this into production, there
were many late nights working and some all nighters (I remember quite a few). In
this picture, we see a bunch of the principles, early in the morning, staring
over Bob Cooper's (I guess, perhaps Bill Aimonetti) shoulder while he watches
the production run. I'm not sure what the letters are on the things that look
like weights. The symbol under the horizontal surface (table) on the screen is
the LINCS (Livermore Interactive Network Communcation System) symbol. Behind Bob
Cooper is Barbara Atkinson. To her left is Jim Minton, in his pyjamas with the
teddy bear. Standing in the doorway, left to right, are me (Jed), Dave Fisher,
both showing signs of wear, and Pat Gray with the smile on his face. I think
perhaps Pat didn't put in the all nighter. This cartoon is also noteable because
it is signed both by BA (Bill Aimonetti) and BC - which I'm guessing is Bob
Cooper. Does that mean that Bob Cooper did some of the drawing? The signature is
This cartoon shows a situation that happened fairly early in
the system testing, I think the date is 10/27/86, when some stress testing was
being done. In the cartoon, Harry Nelson is on the left and I think that's Rob
Haynes on the right. Each has created some 5,000 files without knowing that the
other was doing the same. They are surprised when they are unable to create any
more files, but with the creations by both of them the file system is genuinely
full. The signature is BA 10/27/86 (I think - could be 87 but the content
doesn't fit for so late in 1987.
I'm not sure what this one depicts. That looks a bit like Dave
Fisher, so the "qa" might well be a quality assurance test of the process server
(e.g. one of Gail Whitten's QA test runs). This one seems to be unsigned, but
the date is almost surely December of 1986.
This cartoon shows a QAlib (Quality Assurance library) test
about to begin. That's Gail Whitten lighting the cannon. I believe it's Dave
Fisher on the left and Dave Getchell on the right in the blocks, but I'm not
sure. The signature is BA/MP 87?
Early on, in the transition to production, it didn't take much to make the system
unresponsive. A symptom was that the terminal concentrator which would try to
forward terminal input would be unable to get it through. It sent back an
"@Sorry" message when it's buffers were full and it couldn't deliver a terminal
message. That's Harry Nelson at the terminal. The signature is BA 6/20/87.
During early testing of the system, one of the problem areas was
getting asynchronous system status. At that time, such status was returned with
"escape" messages that started with control-e. The simpliest was control-e "s",
for "status". Harry Nelson, at the bottom on the right, was quick to point out
problems in this area. In fact, in general, Harry was a good diagnostic for NLTSS,
especially in the early days before we had more formalized testing like QAlib.
Early in the meetings, to discuss status, we would often turn to Harry to see how
he was doing. This check became knows as the "Harry Meter". I'm not sure who
that is about to type the carriage return. There is no signature or date.
During this time period Harry Nelson was working on a chess
program that was competing on the national stage (Cray Blitz?). He was using the
NLTSS system for running it because he could get it to parallel process under
NLTSS--but not under LTSS at the time. The control-e-ps command was apparently
something to probe a bit deeper about the status of a process. In this cartoon,
we see Harry is miffed when he types control-e-ps to query the state of his
chess program and it ends up smashing the program. The signature seems to be
(under the middle PS) BA 7/11. Probably 1987.
In this cartoon, we see Dave Fisher and the process server
standing up to the worst that Gail Whitten and her QA library could shoot from
their cannon. Gail seems to be checking her watch to see how long things have
run or are taking. The signature is BA 3/30/87.
In this cartoon, we see Dave Fisher behind the desk starting a talk about the process
server. Such a server was a bit unique to NLTSS. I expect that's Barbara
Atkinson to Dave's right at the table. It might be Linnea Cook on the other side
of the table. It's clearly Harry Nelson with the briefcase that bears the LINCS
name and logo. Who the rest are, I'm not sure. Probably Mike Pratt and possibly
Chris Hendrickson also listening. Behind the door, many of the developers are
listening in and cheering Dave on--though I don't recongnize anybody. I'm not
there in my recognizable hat. The signature seems to be BA 8/17/86.
NLTSS followed the LTSS syste, which it was emulating, in having a special user,
User-2, that ran the software for most input/output functions like servicing
remote job entry terminals. This picture depicts User-2 doing his job. This
suggests that some of the User-2 functions were beginning to work by April of
1987. The signature is BA 4/18/87.
In each of the NLTSS status meetings, it became a standard
procedure to ask Harry Nelson for a system status update, the "Harry Meter". In
this cartoon, we see Harry indicating that a production code has monopolized the
system. I don't see a signature.
Octoport was a network gateway for NLTSS to other components within the LINNCS
network. In this cartoon, it appears that something in Octoport has brought down
NLTSS. The symbol over the prostrate figure (Lance Sloan?) is the LINCS logo.
The Octoport is depicted as an Octopus as the Livermore network was generally
referred to as the "Octopus" network. I don't see a signature.
"Model" is the Pascal-based programming language with abstract
data type extensions, originally developed at Los Alamos, that was used for the
systems programming for NLTSS. Nsyslib was the version of the low-level library
that could be used to access system functions that was developed for NLTSS. This
library essentially supported emulation of the LTSS systems functions on top of
NLTSS. I expect that is Barbara Atkinson playing the role of Miss Piggy on the
right. I really don't know who the others are. Nsyslib was developed largely by
the User Systems Division and in the local Fortran dialect, "LRLtran". I don't
understand the reference to "Model Muppets" programming Nsyslib. The signature
is difficult to read. I guess BA 9/86.
This cartoon shows Norm Samuelson "cooking" up the command line interpreter (CLI).
The command line interpreter was a bit like a shell. It was responsible for
allowing logged in users to send "commands" to the system. It supported
execution of user programs and the various status requests, interruption of
programs, etc. The signature is BA 11/5/87. As I recall this was the last
cartoon Bill drew.
This cartoon depicts me (Jed) standing behind the LINCS symbol/logo and fending off
"bad naks" from the network. I don't remember where those "bad" naks were coming
from. The signature appears to be BA 5/17/87.
This cartoon shows a couple of people setting traps for system problems. Roger
Skowlund, behind the rock, is setting a trap for a "Box and ID Disk pattern"
problem. (Box and ID are remote job entry terms, disk pattern is what the disk
data is set to when there is no file containing it so that it can be recognized
if it shows up in a file.) Harry Nelson, behind the tree), is setting a trap for
some other problem. The signature is BA 8/30/87.
In this picture, I (Jed) seem to be talking to the Kiwi system and having some problems.
I have to admit that I don't remember what the "Kiwi" system was. The signature
is BA 4/5/87.
For information about this page, contact us at: firstname.lastname@example.org