The Dawn of Engineering in the Century of Humphrey and Friends
Paleo-Engineering 205
Mars Institute of Technology
New Pittsburgh, South Sagania
4 July 2202
Abstract: This essay is in fulfillment of the requirements for PE 205 toward my Software Technician class B in Hyper Spectral Systems Design annual re-certification. This Paleo-Engineering study taught me a lot about the origins of the scientific approach to software that we take for granted today. The struggle of our foreparents are cast in the techno-cultural context of the late 20th century showing that software was already vital to them. Their achievements in spite of a dire lack of software science were quite remarkable.
At first, I found the subject irrelevant. Good grief, I can see the environment commissioner’s face if she learned that even one of our biogenic synthesizers software was hacked on a 21st century kernel implemented in C#** by a moonian horde of dilettantes![1] Well, holonet research uncovered fascinating details that made me think. For instance, I learned that the scientific approach to programming that we take for granted today, really originated on Earth in the work of one W. E. Deming in the first half of the 20th century. This Deming seems to have been the Watts Humphrey of hardware production in his days, and his leadership-driven scientific approach can be seen as a harbinger of Humphrey’s work. Besides, he is reported to have done in 20th century Japan what Humphrey later did with more lasting success in the United Asia Pacific States.
I come out of this assignment with a heightened respect for our foreparents. They could not even have imagined our virtual funds network, Sol cometoid defense system, the Federation space traffic control, Mars environment control center, or even our personal health maintenance appliances. But they produced the small-scale equivalent of these software intensive systems through heroic intellectual effort. I can’t imagine our generation developing the aqua forming systems of Europa, our entertainment holonets, the Federation distributed self-government software or practically anything in our society in the way they worked back then. Even our school’s engineering environment for DaD-2195, which is restricted to practice what we learn on the intra-holonet, is light years ahead of how they developed software in the century of Humphrey and friends. What they achieved back then made me think of the Roman Empire.
On Earth, quite elaborate and elegant albeit massive constructions have survived their builders as well as 3 world wars for over 2,200 years. Some still standing pieces can even be seen today south of the vitrified desert of Western Europe. The Romans built impressive aqueducts for instance, without the knowledge of calculus, static or elasticity modulus. De nobis fabula narratur,[2] to put it in their dead language. Well, our foreparents managed to produce elaborate and elegant albeit massive programs on unreliable systems using only their wits and heroic discipline. They did this without architectural calculus, brainstormers, the Belstein[3] compendium, holonet inspections, system synthesizers, quantum computers, etc.
Most importantly, I came to realize that it is our school’s mentoring system and multi-disciplinary sol-wide process that make true engineering possible. It is this process that guarantees the consistent application of the scientific method in our daily work. It is quite striking to me that back in the 20th century, they were able to practice a significant subset of all this. For instance, Humphrey had developed an infant version of our virtual teams system process. His brilliant use of personal and team data, however limited, allowed his disciples to practice Roman-style engineering. What took the rest of them so long? Of course, Earth then did not have the horrendous engineering tort laws our century is famous for. Consequently, the motivation for building error free systems was largely lacking.
Incredibly, most of the programmers of that period were not even schooled in the science of their own time. Today, they would be thrown in some neuro-linguistic reprogramming clinic for illegal exercise of an engineering discipline! It is quite remarkable that humanity came to this planet in 2026 riding on systems patched together by non-certified programmers. Their personal discipline, hard teamwork, and constant scrutiny of the little data they had must have carried the best of them through.
As for the rest, the laggards who resisted discipline and science nearly all the way to modern times, it was not entirely their fault. Their schools, the culture of their companies, feudal management and the very fabric of their society encouraged obscurantism.[4] As difficult as it is to understand today, scientists and engineers were not appreciated in the 20th century. Short-term tactics made them subject to hard labor, education denial, and arbitrary layoffs. This led some of them to anti social behavior such as infecting their infant holonet with crude viruses. Sports figures, 2-D movie stars, lawyers (already!) and virtual funds shufflers were at the top of society back then. Organizations had no psycho-legal ways to attract, reward, and nurture qualified engineers. Some psychologists were working on the problem though. In particular, the name of one Bill Curtis shows up repeatedly in the records that have survived the war.
In other words, the socio economic context of the times delayed the establishment of a true scientific foundation of software engineering.
Version 10A of my concept correlator found a well-preserved artifact on that subject from one Mary Shaw, an influential philosopher of the period. She identified five increasing levels of maturity in any engineering discipline, that finally lead to the scientific method. The holographic image I found is reproduced in Hol. 1.
Back then, they already had grammar theory and various algebraic dialects. The “clean room” method, attributed to one Harlan Mills, had laid the foundation for mathematical programming.
But comparing this to the regimen we engineers have to go through for our certification would be to compare basic arithmetic to tensor algebra. For lack of scientific substance, inconsistent practice, and insufficient use of data to close the feedback loop, mainstream 20th century techniques barely qualified for level 3 on Shaw’s model.
Hol. 1: The Maturation of Engineering
Studying the struggle of our foreparents made me feel a sense of kinship with them. I came to realize that in spite of cognitopharmacology,[5] genetic engineering, and continuous culturation programs, the human condition has not changed. It probably never will. Chopin composed on a piano, Xenastein[6] programs multi-sensing stimulators, but we are no different. We learn, dream, struggle, create, and love like they did; only better.
Sadly, it is the war that unleashed the evolution of
software engineering as we know it today. The brilliant algorithmic
counterattack by the alliance after the sand worm bacterial agent sneak attack
ushered in the era of the cybercorp with its battalions of highly educated
software scientists and engineers. As a belated tribute to the work by late
20th century visionaries from the then “U.S. Space and Air Force”, the alliance
maintained software superiority for the entire conflict. Its elite had been
well prepared to “rely on overwhelming technological leadership and an ability
to respond quickly to the demands of a rapidly changing world”.[7] The endless cycles of cyberwarfare measures
and counter-measures continue to this day. Mars is best at it. Should we ever
decide to dump a tea shipment from Earth in the New Boston spaceport, there is
no doubt that Mars would quickly be winning with software.[8]
After all, de nobis fabula narratur.
Strangely however, even decades after the war, the impact of software engineering on the Earth’s economy was poorly understood by the majority. In spite of the clearest of evidence from the war effort, in spite of brilliant work by econogineering[9] pioneers such as Barry Boehm a century before, the Americas Union leadership kept ignoring blatant statistics showing decreasing student enrollments in engineering and science correlated with the dwindling competitiveness of their states. The decline of this great civilization can be traced to this complacent attitude about their own data. De nobis fabula narratur. Let’s make sure it does not happen to us, here on Mars.
Another connection was suggested by the release 16.25 of the school brainstormer.[10] Back in the mid 20th century, electronics engineers patched together quite elaborate and elegant albeit massive semiconductors one by one to build slow and awkward components for Von-Neuman calculators. However, the components were not standardized, their interface rules were not established, and entire systems were developed in ad hoc manner around the Ale replicator or then equivalent machine. In mid-century, a company from the Americas Union proposed a family of standard components called the “7400 digital ICs series” which resulted in the first microelectronics revolution.[11] De nobis fabula narratur. It was not before the Belstein compendium of software archetypes was assembled in the middle of the last century, that we, software engineers and technicians of all grades and specialties, were able to benefit from certified error free components. Here again, the analogy with the dawn of the electronics age is striking. The Belstein only offered what could be compared to 20th century era Small to Medium Scale Integrated Circuits (SSI/MSI) components. In the following decades, Navyug Musharraf,[12] and then Uhuru Du Motier,[13] Jin Yun Almeraz,[14] and others added more generic and higher level archetypes that can be compared to the LSI, VLSI, ELSI, ULSI evolution of optoelectronics systems of the 20th and 21st centuries. Furthermore, electronic design automated tools of the period with their libraries of components, rules of integration, simulators, and subsystem generators can be seen as the ancestors of our modern system synthesizers. A major difference, of course, is that our systems involve mostly software design work.
Another set of ideas uncovered on the holonet may have greatly contributed to the dawn of software engineering two centuries ago. Victor Basili and his team working at a precursor of our Federation Space Organization, called this set of ideas the “experience factory”. Holograph 2 depicts the basic concepts.
In the experience factory, every project was used as an experiment to test the impact of any new technique or tool. Software processes, components, data, tools, and relevant previous experience were examined and tailored for the project at hand. During the project, product and process metrics were carefully collected. At project end, the impact of the new technique or tool as well as data and lessons learned were packaged into the experience base in forms of models, theories and training for later reuse and continuous improvement. This was a valiant attempt at implementing Shaw’s level 4. Of course, it is also the essence of our school’s basic curriculum which can be traced all the way back to Humphrey’s grand unified process[15].
Hol. 2: The Experience Factory
A routine run of a draft of this essay through the holonet inspection process yielded yet another line of thoughts that may have contributed to the maturation of systems engineering during the period. Someone from the Weinberg Institute in the Barabara colony of Titan[16] pointed out that the century of Humphrey and friends began with Frederick Taylor and the myth of the omniscient guru, the expert/hero, fighting fires one at a time. In the 1950’s, participative management and Douglas McGregor’s Theory Y made everyone fight the fires in teams. The next development occurred when experts started looking at the entire system to prevent fires from occurring in the first place. This was exemplified by Kurt Lewin’s learning organization. The next step, taken in early 21st century was the collective study and resolution of what we now call metasystem problems, taking into account their technological, procedural and psycho-sociological aspects. Imperial management was on the way out. Multidisciplinary empowered virtual teams from all walks of engineering have been working on metasystem analysis ever since.
Hol. 3: From Taylor to McGregor and Lewin
It is this convergence of ideas that lit up the dawn of software engineering over 200 years ago. From Taylor, our founding parents learned of the need for scientific management based on data and discipline. From McGregor they learned about the importance of everyone’s involvement toward clear common goals. Through Lewin’s learning organizations, they demonstrated that People do best by learning and learn best by doing. Humphrey and his friends put it all together. Better than most they understood that to master a complex system, you have to work within it, experience its weaknesses and personally commit to your individual share of the needed improvement.
Without this convergence of ideas, we would still only be fighting what another influential philosopher of the period, Frederick Brooks, called the accidental complexity of software engineering.[17] Without this convergence of ideas, we could never have reached the point where today, scientists, engineers and technicians from all parts of the federation are cooperating in the continuous improvement of the sciences and technologies that drive the progress of our respective disciplines. Here truly lies the enduring legacy from the times when Humphrey and his friends were inventing the future.[18]
Credits: Many thanks to Rosie Wood (Aimware) for her review and excellent suggestions