The Weight of Software

A story, from Jeff Silverman:

Whenever you build an airplane, you have to make sure that each part weighs no more than allocated by the designers, and you have to control where the weight it located to keep the center of gravity with limits. So there is an organization called weights which tracks that.
For the 747-100, one of the configuration items was the software for the navigation computer. In those days (mid-1960s), the concept of software was not widely understood. The weight of the software was 0. The weights people didn’t understand this so they sent a guy to the software group to understand this. The software people tried mightily to explain that the software was weightless, and the weights guy eventually went away, dubious.
The weights guy comes back a few days later with a box of punch cards (if you don’t know what a punch card is, e-mail me and I will explain). The box weighed about 15 pounds. The weights guy said “This box contains software”. The software guys inspected the cards and it was, in fact, a computer program. “See?”, the wights guy said, “This box weighs about 15 pounds”. “You don’t understand”, the software guys responded, “The software is in the holes”.

16 Replies to “The Weight of Software”

  1. This kind of story seems to be one of those things that keep happening, and while I can’t speak for this particular story, they really do happen. You have some bureaucrat that doesn’t quite get it about science or technology, who then proceeds to do things like blindly follow a checklist or insists that what they have been told _has_ to be wrong somehow.
    I tend to call them vacuum pocket stories, after this incident:
    http://www.sciforums.com/vacuum-pockets-and-safety-nazis-t-41446.html

  2. I think this is a slightly “stolen” story. Two or three years ago, while listening to Science Friday on NPR, the men who worked in the labs that developed the first GUI, precursor to the internet, and other ground-breaking computer related innovations, were interviewed about their lives, work, awards, and thoughts on the current state of technology. One or two of them related this story in the setting of working on a military aircraft project, with the role of the suspicious character played by, if I remember correctly, a non-com who had sneaked into their “office” over a weekend.

  3. Interesting view, but flawed: The weight of the software is not in the holes, but in what the holes imply.
    Which is to say, software cannot run on ether. It needs hardware, and how it’s written has a bearing on what hardware it needs.
    So yes, software does “have” weight.

  4. Oh yes, these things happen routinely. I say this from my experience on Army, Navy, and Air Force contracts at Boeing, Lockheed, Rockwell, and their kin.
    When they offered me the position of Software Manager on the Air Launched Cruise Missile, which I politely declined, I’d been in a meeting where I seemed to be the only actual scientist or engineer in a room filled with clueless managers.
    They had some concerns about the aerodynamic stability of the ALCM after it would have a fissionable warhead. I’m being intentionally vague here, for security reasons.
    I said that I could rephrase the question, and did so thus.
    “You have a legitimate uncertainty about the distance between the center of pressure and the center of mass when a different payload is retrofit, right?”
    A suit from Aero agreed.
    “Well, then,” I said. “I have a solution, representing my Software Engineering Department. Simply insist that the payload use variable-density fissionable material. Then under software control, the density can be increased or decreased until the distance between the center of pressure and the center of mass reaches precisely the specified value.”
    They all agreed, showed relief and gratitude, and broke for coffee. Several congratulated me over coffee.
    The next day I declined the management position.
    I am not going into the incident with the fraudulent “nuclear battery” (a fraud that made the Wall Street Journal). My colleague Ed McCullough listened to a couple of hours of nonsense at the meeting where a proven plagiarist who had failed to graduate from community college pitched installation of the “nuclear battery” on the Space Shuttle. Finally Ed spoke up.
    “I was in charge of the nuclear reactor on a Navy submarine. Is there anyone here, besides myself, who has ever taken even a single class in nuclear chemistry, nuclear engineering, or nuclear physics? No? Then I politely suggest that this meeting has been a complete and utter waste of time, as none of you know what you’re talking about.”
    And don’t get me started on the hundreds of thousands of dollars my division spent in its “black budget” on an “engine that doesn’t need to be refueled” — a version of the water to hydrogen and oxygen gas to water scam that dates to the Civil War. Both Ed and I were called to testify at one of the lawsuits on that perpetual motion machine.
    Oh yeah, these things happen routinely. I still have over 7,000 pages of my “Pearl Harbor File” designed to protect me if any of these things became dangerous. Even that strategy failed, but it would take too long to explain.
    Ed McCullough is still at Rockwell, or technically the part of Rockwell that got acquired by Boeing, having become a manager himself, and a big wheel.
    Edward McCullough came to Rockwell with mining experience: he had been in charge of the world’s northernmost mining site. His mentor at Rockwell was a consultant to the mining industry. The Lunar/Planetary group at Rockwell was arguably the best in the world, when I was in it, until they elevated a plagiarist PR flack into it, and fired or transferred away everyone from MIT and Caltech who made the creep look bad.
    Edward McCullough is a principal scientist at Boeing. He has received his professional schooling mainly in nuclear engineering through the U.S. Navy (gaining his Certification for Nuclear Engineering in 1975). Mr. McCullough focuses on concept development and advanced technology at Rockwell Space System’s Advanced Engineering and Boeing’s Phantom Works. He has researched innovative methods to reduce the development time of technologies and systems from 10 to 20 years down to 5 years. He has experienced successes in the area of chemistry and chemical engineering for extraterrestrial processing and photonics for vehicle management systems and communications. This included leading a chemical process development research team in a skunk works environment for 4 years. Mr. McCullough has led efforts for biologically inspired multi parallax geometric situational awareness for advanced autonomous mobility and space manufacturing. He recently developed several patents, including patents for an angular sensing system; a method for enhancing digestion reaction rates of chemical systems; and a system for mechanically stabilizing a bed of particulate media. Mr. McCullough has served in a variety of professional societies and councils. He is a member of the Board of Trustees for the University Space Research Association, a member of the Science Council for Research Institute for Advanced Computer Science, and a charter member of the AIAA Space Exploration Program Committee.
    Of course, I can only testify about the insanity in my era of those projects. Ed can go on for hours about how it still goes on — and how he learned to avoid the craziest anti-scientific loons in the business.
    Your tax dollars at work.

  5. From a couple decades as an engineer in private industry I think it’s well known that about 10% of the engineers do 90% of the actual work. The vast majority are restricted to making inane comments and holding coffee mugs.
    The basic problem is that it is difficult to measure what engineers do. Suppose you screw up your assignment really badly. You end up delaying the whole project. The result is that you have to work late into the night for a couple weeks. Management concludes that you were assigned the most difficult part of the project and performed heroically. Or better, they conclude that your good work attitude makes you the perfect candidate for promotion to management, LOL.

  6. Much truth in what Carl Brannen writes.
    On a successful project, if you ask the N people in the project {p_1, p_2, p_3, …, p_N} what percentage of the work they, personally, did, and add up those percentages, the total number is typically over 200%.
    Success, in management, is often achieved by failing at bigger and bigger projects worth more and more money.
    Sure, you lost the bid for the $10^9 project, but you did spend $10^8 on the proposal and managed several times 10^3 people, so you’re obviously qualified for Executive Management.
    Don’t fail down. Fail up.

  7. Of course, one can weight holes. In high school, I got a sample of indium antimonide from the head of IBM’s Thomas J. Watson labs and did Hall effect measurements. From them I could determine the mass of the carge carriers. That mass was negative, consistent with the material being a p-type semiconductor.
    But what is the mass of the holes in the punchcards, to a nascent Homer Simpson eating Donut holes?
    And does entropy weigh less than enthalpy because it is spelled with 1/8 fewer letters?

  8. The theorertical minimum weight of software is
    m = k ln(X) t / c^2
    where
    k – Boltzman’s constant
    ln – natural lograithm
    X – number of ways it might have been written. For a 10k source file, this is 2^81920
    t – max temperature, in kelvin, that the substrate on which the software is written can handle without bursting into flames
    c – speed of light. Natch.

  9. True Dave, but since information is only useful if you can use it somehow, it must have a carrier or processor at some point, so…..how about USEFUL information is physical?
    This gets into semantics a bit
    Paul is right (which gives you about a mass/c^2 of 1.5 eV for the 10k source), if I erase intermediate results.
    But I guess it will “weigh” more if I include error correction bits?
    So does mass result from erasure of information?????
    And does this mean that Linux literally weighs less than Vista????

  10. to take this a bit further,
    mass comes only from irreversible processes, which can be modeled by non-Hermitian Hamiltonians which lead to non-unitary dynamics, so mass comes from non-unitary processes such as might arise in the loss of information as a black hole evaporates……
    Could be a good sci-fi story

  11. I’d be inclined to argue that the weight of the software was negative – “no software, no holes” :-).
    This also reminds me of a Dilbert thread (sorry haven’t been able to dig up the reference) from a few years back, where the question was implicitly raised by the PHB, as to whether a 0 or a 1 recorded on a hard disk, is heavier.
    From the discussion thread I saw, it turns out that the 1s may actually be heavier. Though, as you’d expect, not by much.

  12. Information is only defined with respect to something else… It is how much the uncertainty about the state of object A is reduced by knowing the state of object B (B would be the media ‘containing the information’ in this case). Explaining how code is information in this sense gets quite confusing though, even for CS types.
    Yeah, there is a theoretical minimum amount of mass (or energy) that is required to store a bit of information… I always found that as profound as (or even more than) E=mc^2
    BTW: The story Rick linked about ‘vacuum pockets’ is great…
    PS: In this case, c is a fundamental physical constant. It is a bit confusing since the speed of light in a vacuum happens to equal c 😉

Leave a Reply

Your email address will not be published. Required fields are marked *