The last section told a story of a strange social phenomenon: MIT was split down the middle, containing two cultures deeply alienated from each other. Their worldviews were sharply at odds. Yet, curiously, their opinions lived on different planes and were expressed in different modes. So the differences were ignorable. The job of reconciling the two worldviews was orphaned. Each side was of the vehement opinion that the task was the other sides’ problem. Like Douglas Adams’s fictional spaceship, the problem was fully enclosed in a Somebody Else’s Problem field. It was rendered effectively invisible.
To my amazement, this state of affairs has been stable for twenty years. We see the effects on our politics today. In a professorial mode, Elizabeth Warren patiently explains the theory of the problems of the tech industry, and suggests corresponding regulation. Tech pundits immediately protest in unison: “Inappropriate Regulation!” Referencing an unstated shared worldview, they pan Warren’s efforts. But they don’t offer an alternative diagnosis of a problem. Nor do they suggest regulation they would consider more appropriate. They don’t seem to consider that their problem. Nor does Warren seem to consider it her problem to dig out an explication of the unstated shared worldview. So the impasse remains unresolved.
As we said, we are going to organize our story like a road-trip movie. Our characters will fight, and make up, and fight some more. This is the first section owned by the Pscyhohistorian. He has been paralyzed by the coconut warriors blowdart. We are forcing him to talk to the engineers, whether he wants to or not. We are forcing him to face that their worldview is not in a clean graduate-seminar-ready state. And dealing with it is his problem, not ours.
In this section, his counterpart is an ordinary engineer. It represents my state in 1998, the year Krugman first wrote about Microsoft. (I refer, in this dialog, to some of the things I was doing that year. It reminds the reader that there was a time I did things besides prognostication. This work flowed from the energy of that previous effort. )
The Rumble Is On!
But anyway, enough preamble! The rumble is on!
Psychohistorian: Krugman said he would listen to anyone who could “state their assumptions,” whether they were professionals or not. That should give you an idea why he isn’t listening to you. You’re falling flat on your face in the “statement of assumptions” step.
Observations…
Me: But why are you insisting we have assumptions? Fred Brooks did not say “there must be upward movement” because he derived the result from first principles. His Fish Ladder model was not constructed: it was discovered. It was an observation, a distillation of experience from untold numbers of projects he had seen go well or badly. It caught on because it resonated with untold numbers of others with similar experience. Are you saying such observations are ignorable just because they aren’t embedded in an appropriate derivation? It’s a pretty big deal to blow off Fred Brooks.
Psychohistorian: Calm down! I’m not blowing him off. I’m listening, aren’t I? But I’m puzzled. You say that this statement is merely an observation, not a derivation. But observations with no context are difficult to understand. At another point, you quoted Stephenson pointing out the device drivers Palm or Lego contribute back to Microsoft. OK, I’m looking. And I’m asking myself: so what? After all, Krugman has many things to look at: trade wars, the problems with in Europe, the radicalization of the Republican party. And you would say: drop all that and look at the device drivers swimming upstream? Wouldn’t it be quite fair for him to ask: um, why? It hardly seems important next to all those more pressing concerns. So maybe we don’t need that observation derived, per se. But we need some idea why it matters.
Not On The Map…
Me: Well, Krugman was angry because his cost-benefit analysis was dismissed. And since I sympathize with his unhappiness, I might wish to take up the challenge. But doing so, I am immediately confronted with a problem: nothing in his picture seems real to me. I mean, if we unwind the abstraction, the two Barons on the river are separately selling an operating system and a browser. We might ask: where are the real-world Barons who hike up the price of browsers the moment their river traffic is uncoupled? I don’t see any. Browsers are free and always will be. So, um, price benefit analysis? What? How? I mean, I’m sorry I’m so stuck on this idea of actually observing the real world. But it is hard for me to get past the unreality of the picture. If the rules of your game don’t allow mere observations, what else am I supposed to say? As you said, these observations may be too bizarre to be of any use to a policymaker. But maybe a scientist should pay attention. A model’s match with reality matters. Doing better science is a reason to care.
Psychohistorian: I agree that observation is important. And we are trying our best to be good scientists. But all this is much more problematic than you make it out to be. You may not be fully aware of everything you are doing when you “observe.” When you say “Browsers are free” and then blithely add “and always will be” you aren’t only making an observation. You are also asserting an internal dynamics which keeps the observation stable—and makes it something other than a fluke. You understand that dynamic. But you aren’t explaining the understanding. You might not be even quite aware of it. Your understanding is there. It isn’t a “mere observation.” But, as Krugman would say, you are sleepwalking.
Psychohistorian, continuing: After all, there are many observations one could make, but only a few of them are powerful and predictive. For instance, a reaction to the fisherman pointing out the red sky might be to ask “OK, the sky is red now, but sometimes it is yellow, sometimes blue. What’s the big deal about red?” The point is: all the sky colors are observations. But some have much more weight than others. Only a few are key to understanding the logic of the world. It was a particular color at a particular time of day that had predictive power; all the rest were ignorable. And the fisherman knew that. His experience caused his eyes to pick out the only observation that was powerful and predictive. For him to then say he was “only making an observation” would be disingenuous. It was the act of singling out that one observation over all others which had content. And he was omitting the logic that guided this crucial insight. That’s what we mean by “sleepwalking.”1 Now, maybe the fisherman was doing it because he didn’t know that logic. The clouds were hidden around the curve of the earth. He didn’t know they were there. But there was logic. The logic was the crucial content of his observations.
Me: Hmm… OK, let me try to pick out a more key observation for you. To tie into Fred Brooks’s observation, there is a deeper problem with the Rhine story. Implicitly, the price benefit analysis implies the barge carries something inert, expensive, heavy and difficult to transport. We imagine steel or coal or other industrial products. But if I think of code as a thing on a river, I don’t think of it as inert. I’d picture it as light and active to the point of being self-powered—like salmon who sometimes can climb a fish ladder.
Are Only Concrete.
Me, continuing: After all, there is a reason luminaries like Tim Berners-Lee spend so much energy describing the architecture of the World Wide Web; and David Clark of the Internet; and Fred Brooks of his hoary old ’80s-era systems. They are struggling with the problem of communicating to managers and politicians. These engineers know the engineering details of the systems very well. But when describing it a problem of management or politics, they struggle. It is painful. But Krugman doesn’t seem to notice that struggle—he ignores it completely. He overwrites their efforts with his own answer: to him, their code is a hunk of steel or coal in a barge. He claims the only thing that matters about it is how much it costs. That answer is wrong. I don’t know what the right answer is, but that answer is absurdly wrong. It is not just wrong, but its blithe self-assured wrongness is offensive. What right does he have to be so arrogant to assert he knows so much when he knows nothing?
Psychohistorian: If you are reading this story as arrogance, you aren’t understanding its purpose. It wasn’t meant to be an assertion of knowledge of the world, so much as the starting point of a discussion. Obviously he knows a lot less about the Web than Tim Berners-Lee; a lot less about networks than David Clark; and a lot less about large operating systems than Fred Brooks. But he does know one thing none of them do: how to fully explicate an argument. He knows how to articulate the logic from beginning to end. That is supposed to be helpful.
Psychohistorian, continuing: Because it is so completely articulated, if you don’t like where it ended up, you are invited to trace back to the point where it went off the rails. He is clearly shocked no one took up the challenge. It is not meant to be hard! Whether or not you think it is a good model of your world, it is the standard model. So he wants to know whether you agree with it. If you don’t agree, he wants to know where your understanding diverges—and why. The only answer he’s getting is a mush of “this changes everything” and inarticulate antitrust arguments. He has reason to be frustrated. I’m sure he never expected his audience to be having such difficulty with the task. He certainly didn’t expect them to become so furiously alienated! It is an easy challenge.
Everything Is Strange…
Me: But it isn’t easy! Nothing in my world is on his map. Nothing! Think of all the features in Fred Brooks’s story: the salmon jumping the fish ladder, the Borg ingesting them, the beasts stuck in the tar pit, the nightmare werewolf for which there exists no silver bullet. Where are those things on Krugman’s map? Nowhere! It’s all different. How can I line up two threads of logic when nothing corresponds! How can I address listeners who ignore everything that isn’t on their map—when nothing is on the map! Everything I see is strange. It is not only strange but strange with a vengeance. My eyes pick out not the River Niger, but the men with mouths on their stomach!2 In order for there to be a point of divergence there has to be a correspondence first. And there isn’t any!
Psychohistorian: Oh, come on, that’s not true! You have a river, after all. I mean, Fred Brooks’s replacement of the Waterfall with the Fish Ladder did at least allow the notion that a river is a useful abstraction. OK, this may be shockingly basic. But since we have so few correspondences, let us take whatever we can get. You agree that there is a supplier and a consumer—a consumer who might in turn use what they receive from the supplier to produce another product. In that way the product moves continuously, as down a river. You are agreeing that it is useful to think about their relationship in a canonical, abstracted way. You disagree about the nature of the canonical relationship. That’s the conflict between the Waterfall and the Fish Ladder. But in this you do at least have a point of correspondence and a point of departure. Don’t freak out! We’re getting somewhere. You can do this!
But Logically Connected…
Psychohistorian: Not only that, in this framework it becomes apparent your seemingly phantasmagorical observations are logically connected. Someone who is trying to arrange the things in their basement will naturally spend energy checking and rearranging the labels on the boxes. Said this way, it is quite intuitive. But if we cast it into the less intuitive river picture, we get exactly the puzzling leaping salmon made famous by Brooks. The homeowner naturally wants to go down to the basement to check the labels on the boxes. He wants it much more than he would want to go out with clippers to touch up his scraggly lawn. Overlay onto this picture a formal model of a flow from producer to consumer. We get energy moving “backwards” from consumer to producer, as asserted by the Fish Ladder model. So there’s an explanation of your active salmon: it follows from the importance of encapsulation.
Psychohistorian, continuing: So you do have a logical argument: you have assumptions that lead to conclusions in a natural order. Labeled boxes implies leaping salmon. Leaping salmon imply fearsome Borg. The search for the silver bullet leads to the labeled boxes. Fear of the werewolf leads to the search for the silver bullet. It isn’t the unstructured menagerie of phantasmagorical creatures you are making it out to be. You just aren’t systematically connecting the assertions in their natural logical order.
Note: This exchange diverges dramatically from the real dialog. It is vastly cleaned up and simplified. I’m mindful that my reader still has to be convinced that I am capable of making sense. So I set up the hairball carefully to make it easy to comb.3 In reality I didn’t immediately see the correspondence of the two rivers, nor immediately grok the “river” way of thinking. Nor did I see that “encapsulation” implies “leaping salmon”. Incidentally, if this lightning-fast exposition is difficult to follow, don’t worry: we will go over this more systematically later. Right now we are digging backwards through the chain of logic to pull out our unstated assumption. Once we get to it we will switch direction to go the more normal forward direction. That restatement will be more slow and methodical; perhaps even slow to a fault.4
With An Assumption…
Psychohistorian: So, now we’ve sorted out an order to these assertions, we can return to our original question: what is your assumption?
Me: OK, you win. You’ve convinced me the observation about “encapsulation” is my beginning point. That’s my assumption. Everything else is a derived result. Uh, now what? Am I done yet?
Psychohistorian: Nope, not done. Having made that observation, you need to give it a name. You have to find it on the map. You have to state how it is a point of divergence from the standard model.
Me, frustrated: It’s still not on the map! It still has nothing at all to do with the standard model! And it’s not my fault! Krugman even noticed it was important, but he didn’t place it on his map either! If he didn’t do it, why do I have to? Argh! Aaaargh! Aaaaaaaaargh!
That Is…
Psychohistorian: Calm down. Calm down. There’s a reason I have to keep torturing you. Remember: the real content in folk wisdom is not the features that your eyes pick out. It is the reason why you pick them out. That knowledge does exist in your head—even if your somnambulism makes you unaware of it! You know it much better, and with much more certainty than Krugman does. Krugman can make the observation. But only you know why it is so central. Only you know the consequences. That’s why I have to torture you.
Psychohistorian, continuing: The burning question is: of all the things you could have chosen to look at, why are you looking at this principle of encapsulation more intently than anything else? Why are you not only looking at it, but apparently spinning out a whole worldview’s worth of further derived consequences as well? Why are you so certain of this new worldview? Why do you feel free to blithely assert every element of the standard model is irrelevant in your world? That’s shocking, isn’t it? You just dumped the standard model? Why are you doing all this without even thinking it all radical… or even surprising?
Me: Why? I didn’t dump the standard model to be radical. It just isn’t useful. It doesn’t describe my world. Engineers need to get real work done in the real world. They can’t afford to live in a fantasy, no matter how “standard” it may be.
Me, continuing: And to your other question, why shouldn’t I look at the principal of encapsulation? Why shouldn’t I care? As an aspiring programming language researcher,5 we live and breathe the dream of better encapsulation. Usually we are dealing with many levels of refinements. However, in some underlying sense it is the only thing we think about from morning to night. Then we wake up again the next day to think about some more! After all, just look at what Fred Brooks said: he came across his principle after a twenty-year hunt for a “silver bullet.” He was desperately looking to find “even one order-of-magnitude improvement in productivity, in reliability, in simplicity” of software production. He was looking for way to slay the problem of projects that “transform unexpectedly from the familiar into horrors.” You don’t understand how true that is! How easily one can get hopelessly lost in a sea of complexity! How quickly badly built interfaces can eat all your time, your effort, your life! Sooner than you can imagine you are doing nothing but unsnarling yourself from their tangle. Look at me right now: I wanted to illuminate deep principles of mathematics.6 In practice, instead, I am suffering over incompatibilities of Javascript implementations. Oh, it hurts!7 Do you want me suffering like this? Do you want to see this horror eating my time, my effort, my life?
Still Unstated…
Psychohistorian: Wait, wait, wait… I have an idea. “Your time, your effort…” Wait… I get it now! We are such idiots!
Me: What? Huh?
Psychohistorian: Don’t you see? What you just said. It’s on the map. Not only on the map, but very on the map. It is what everyone should be expecting to see. What we should be looking for even! I can’t believe what idiots we are for not seeing this right away.
Me: Really? Um… why? I don’t get it. Doesn’t look to be on the map to me. Where? What? What are you seeing?
Psychohistorian: Listen to yourself! What were you just yelling at me about so passionately? Your time, your effort, your life… that’s what this is all about for you. In other words, your productivity. And it depends on those interfaces; those labeled boxes; those compatible or incompatible Javascript implementations. The central force that transforms your worldview is the connection between your productivity and a thing that enables or destroys it. In Brooks’ telling, the werewolf and the silver bullet that slays it. There is a standard name for such a thing. It’s on the map. It’s very on the map. It is quintessentially on the map.
Me: What? Huh??
Psychohistorian, continuing: Have you ever, by chance, heard of the idea of “a factor of production”?
Me: Um, maybe somewhere… isn’t it a Marxist thing? One of those elements of the discredited communist creed? We’re staid engineers—we don’t go in for that kind of thing.
Psychohistorian: No, silly! OK, maybe Marx used the idea, or misused it. But that’s not where it came from originally. It is the definition of a new economy. A new economy is a change in the principal factor of production. In other words, a change in the thing that makes people productive. And that’s what you are telling me: that you care so much about “encapsulation” because it is the thing that makes you productive. And it is different than the factor of production in the standard picture. And as a result of that one change, everything else changes for you. That it is not on the map is the point. That’s not strange. It is the expected result of a transition to new economy.
Me: Really? I thought a new economy was some big new technology that “changes everything,” like electrification.
Psychohistorian: No! Despite rumors to the contrary, a new economy is not a fancy new product or technology or any of that. I know you guys prefer to crow about such things. But that’s not what it is.
Yet Expected.
Psychohistorian, continuing: Now listen to what you are saying: that there is only one thing you care about. You care about it overwhelmingly. What is it? The thing that makes you productive. It is new, not the same as in the standard model. As a result, the rest of the your worldview spins around that fixed point like the stars spinning around Polaris. That is exactly what is supposed to happen in the event of a new economy. It is the expected result. Maybe it is a strange transformation; maybe even terribly strange. But it is an expected strange transformation. So why have we had such trouble seeing it?
Me: But, but… we’re talking here about the productivity of software production. Who cares about that? Well, maybe Fred Brooks cares, and Gerry’s Kids8 care, and the PL9 crowd cares. And other engineers care. But to a first approximation, nobody else cares. It is just us. When Dertouzos was asked about productivity, he wrote a book about industrial productivity. He assumed it was the only thing anyone cared about. Sure, he knew all about Brooks’s werewolf and silver bullet and all that. I’m sure he knew it inside out and upside down. But he didn’t mention it. He assumed Sloanies wouldn’t care. Of course we care overwhelmingly about the productivity of software development. But isn’t it arrogant to assume our private obsessions would matter to anyone else?
Nobody Cares…
Psychohistorian: As you yourself pointed out, don’t make assumptions about whether or not people care! The private obsessions of engineers become everyone else’s problem eventually. Perhaps even, sooner than anyone expects. And if you think the “factor of production” is a question about industrial productivity, you’ve completely misunderstood it. If it were, then the question facing Adam Smith would have been about agricultural productivity. But instead it was about the productivity of factories. That question might also, in the 1770s, have seemed the private obsession of an extremely small band of entrepreneurs. Nobody cared except a bunch of crazy pin makers in the forests of France! There weren’t many factories back then. Everyone cared about agricultural productivity. There was no prospect on the horizon that machinery would help that anytime soon!10 So they too might have thought no one would care. But they would have been wrong. Extremely wrong.
Me: But… you’re talking about all this like it is a huge, world-changing revolution. I never said it was. All the noise about revolution is the work of marketers. Actual engineers don’t talk about “changing everything.” Perhaps we don’t even believe anything has changed. Did Fred Brooks talk of revolution? Of course not. Instead, he said “it is always thus.” All he talked about was the problem of good software engineering. He talked about how hard it was. He talked about the discipline and cleanliness necessary to make progress. He didn’t talk of revolution.
Me, continuing: We’re just trying to make better software, and make software better. We don’t want to fuss with revolution. Too much bother. When I gave a talk on this subject at CSAIL, the first reaction was a professor in the front row who yelled at me: “don’t distract our graduate students!” We just want to get our research done.
About Revolution…
Psychohistorian: Wait… you’re telling me that you are totally re-envisioning the terms of the relationship between consumer and supplier, that you have created for yourself an entirely new bogeyman in the form of your Borg monster… but you don’t want to change anything? All that isn’t a revolution?! If that isn’t a revolution, what’s your idea of what a revolution would be?
Me: Um… well… I guess all that does kinda constitute, um, well, um, some amount of change. We are just too used to it. Remember, we are like fisherman out in the wind and weather everyday. In such a position, even if it rains frogs, frogs seem ordinary. If experts get too surprised by rumors of amphibian precipitation, one just stops talking to experts. No matter how strange it all is. Now that you mention it, maybe it is rather… extraordinary.11
That Technology Wants…
Me, suddenly anxious: Also, um, if you talk to real engineers, please, please don’t spazz out about how strange it is. It will just make everyone uncomfortable. Everyone’s already pretty alienated. Whether or not it is a reasonable reaction, making a big fuss would just alienate people more. It isn’t that we want that kind of change. It is that technology wants it. We’re just going along with what it wants. Don’t look at us! It isn’t us. Perhaps we actively don’t want change. Too uncomfortable. Too disruptive to real work. For instance, Linus Torvalds called himself an “Accidental Revolutionary.” It all seems like a surprising accident.
Psychohistorian: But as you yourself pointed out, the truly huge transformations in history were heralded by exactly that phenomenon. Revolutionaries who want revolution don’t do squat. But engineers who insist that they are only reluctantly pushed into activism by “what technology wants”—whose revolutionary predilections are “accidental”—they are the ones who change everything. Or tragically fail to change everything in time to avert disaster.
Me: Do you think we are tragically failing? Tragically failing at what?
Psychohistorian: Well, we have to think about it. What you are telling me is slowly sinking in… In particular, your Borg monster, the one who eats the leaping salmon… I’m wondering what that’s all about. Krugman mistook your animus against the Borg for left-wing radicalism. You foamed at the mouth with fury, protesting it was all driven by a true concern for technology, not anti-corporate ideology.
Me: It hurt that he didn’t believe! Nobody cares two figs whether Bill Gates is or isn’t rich, or did or didn’t control a powerful company. It isn’t motivated by feelings about power or wealth, as Krugman alleged. No: all we care about is that he is threatening the integrity of the Web. There is a dream of the power a free and open web might offer us—if only it was allowed to come into being. My real work right now is trying to port mathematics simulations originally written for SGI workstations into Javascript. We dream of supercomputer-level computing available to everyone, without friction. Gates is threatening to kill the dream. We are going crazy over it.
Psychohistorian: OK, I’m beginning to get an inkling of what you are talking about. And it scares me. Krugman thought you were overstating a threat; we may have been understating it.
Me: Please please don’t mistake all this for left-wing radicalism! Your talk of “factor of production” makes me uncomfortable… it sounds too Marxist, or at least leftist. One reason I’ve agreed to this debate to gain tools to fight back against the allegations of extremism. I’m not sure I’ve won. I hoped a properly articulated abstraction would clear my name. Instead it is only besmirching it in a different way. We’re not radicals, you understand that? Fred Brooks is a Republican, you know. If I even seem to suggest we’re all a bunch of leftists, my friends will kill me.
That Isn’t Radical.
Psychohistorian: OK, OK! Chill! I won’t assume you are radical. After all, the “factor of production” thing was my idea, not yours. Just because left-wing radicals use the idea doesn’t mean it has to be used in the service of radicalism. Or that it even has to pull left. But even if I grant the purity of your motives—or even additionally acknowledge the ardent loyalty of your beating Republican hearts—I’m not so sure I grant the soundness of your logic. Now that I’m beginning to figure out what clouds leave shadows in the shapes of your crazy menagerie of beasties, I might be more convinced of Microsoft’s ’90s-era nefariousness. But I’m less convinced of his status as an evil monopolist of Rockefeller and Morgan vintage. Your account of his villainy seems off… perhaps by a century or more.
Me: It wasn’t our account. Seriously now: have you ever tried to get David Clark to talk about anything except the mind-glazingly boring details of the actual architecture of actual networks? I have: it doesn’t work. He defers to you economists when it comes to abstraction. You listened to him and hallucinated the wrong abstraction. We didn’t do it. I know: I ran around the lab arguing that they only way to break a bad abstraction was with a better one. And no one bit. The enthusiasm just wasn’t there. There’s a reason engineers talk with a shocking specificity. There’s a reason they drag you through the mind-numbing details of middleware and APIs and Internet protocols. They are afraid of abstraction—afraid of its potential to mislead. Maybe they should be; maybe you are not enough afraid.
Psychohistorian: OK, you have a point. Maybe we should be more afraid, maybe we should listen more. But I think we are in trouble. We aren’t going to get out of the mess we are in without wielding the power of abstraction. But perhaps you are right we need to be more careful. To that end, we need to do this logical argument thing right. Stringing along a menagerie of beasties in a rough order of presumed deduction just doesn’t cut the mustard. We need to be a lot more careful.
Start Over Now.
Psychohistorian, continuing: I’m taking over from here. We are going to start this story all over again at the beginning. We are going state every step of our logic completely. Maybe it will take a while; maybe we still have a tremendous amount work to do before we are ready to address the issues of the day. But I have faith doing the work properly and thoroughly will pay off. I am stunned you people are willing to live with such loosely stated logic. Doesn’t it drive you crazy? As you told me Fred Brooks dreamed of “the replacement of demon theories and humors” with a “germ theory.” He thought “progress would be made stepwise, at great effort, and that a persistent, unremitting care would have to be paid to a discipline of cleanliness.” Perhaps if he wanted such a dream of cleanliness, he shouldn’t have addressed other engineers. No, instead, he should have talked to an economist!
- TODO find reference where Krugman complains about “sleepwalkers” ↩︎
- This is a reference to a line from another part of Krugman’s essay about methods and models. He actually has two versions of the parable about how model building induces blindness. I picked the meteorological one for brevity. The other one is a gem; it was only reluctantly cut for space. It talks about how maps of Africa drawn in the 15th century had many interior features. Some were real, like the River Niger. Others were completely false, like “regions inhabited by men with their mouths in their stomachs.” When experts redrew those maps in the 18th century, both the real and the fantastical features disappeared. The interior was empty—both of the fantastical stomach men and the real River Niger. In the dialog, I’m complaining: if my story sounded like the River Niger, maybe I could stand up for it. But since it sounds like men with mouths in their tummies, it is hopeless! ↩︎
- I conveniently picked elements of “folk wisdom” that were merely the third or fourth step in the unstated argument, rather than something ridiculous like the eighth or fourteenth. A reader wanting to see an authentically undigested ball of hair could read Neal Stephenson’s “Command Line.” But don’t expect it to make any sense! Without this careful choice of entry point this whole effort wouldn’t work. I’m telling a short, simplified version to give an idea of the method—while cutting out most of the true agony. For instance, later I will quote my favorite bit from Stephenson—many chapters from now.[TODO:reference] It takes a lot of structure to fit in his observations. To give an idea of how crazily Stephenson jumps around in his argument, note he starts with the assertion that Microsoft villainy is not that of a classic nineteenth monopolist, but something else instead.[TODO:reference] Fourteenth step, indeed! We are carefully leaving such a contentious issue aside. We will hold off[TODO:reference] on that until we have a lot more structure. ↩︎
- This single section of dialog contains the results of a few years of thought. A friend, on hearing I was rewriting, asked “so are you killing your darlings?” ↩︎
- Actually, this is messing with the timeline a bit—I didn’t get seriously into programming languages until well after this time. When this dialog was set I was still transforming from a math and physics undergrad to computer science. I had a dream to make explanatory math animations on the web. I had already done it on SGI supercomputers. But the web beckoned. Though technologically impoverished by comparison, was just so much cooler. Originally I tried using Java, but a friend tipped me off (in the mid ’90s!) that Javascript was the future of web programming. So I was trying to use Javascript… and suffering bitterly. This was the mid ’90s, remember. It was bad. ↩︎
- The math I’m referring to here isn’t the mathematics of economics modeling, but the mathematics of classical mechanics. ↩︎
- Mid ’90s era Javascript, remember. I didn’t have AJAX or React or Node or any of those wonderful new things. All I had were the problems these were meant to solve. Plus incompatible and even broken implementations as well. It hurt. ↩︎
- This is not a reference to the band or the charity, but to the acolytes of Gerald Sussman. He, along with Hal Abelson, wrote “Structure and Interpretation of Computer Program” (SICP for short). It is among the oldest and hoariest books about computer programming. One might characterize it as a more technical continuation of Brooks’ vision. Like Brooks, it preaches sanitation as a bulwark against disease in the dirty world of programming. ↩︎
- PL is short for Programming Languages. One might say “Felleisen’s and Krishnamurthy’s Kids.” But that wouldn’t roll off the tongue in quite the same way. At Northeastern, Felleisen was fighting to bring the SICP dream of sanitation to the masses. ↩︎
- Later[TODO:reference] we will discuss Hamilton’s opinion on the subject in 1791: “It shall be taken for granted, and the truth of the position referred to observation, that manufacturing pursuits are susceptible in a greater degree of the application of machinery, than those of Agriculture.” For that reason, he argued for support of manufacturing with an apologetic air. ↩︎
- By the way, this is another point where I am oversimplifying. In reality at first I flatly, stubbornly refused to believe the implication of revolution. It just didn’t sound the way I felt about the situation. Like Fred Brooks, I was deeply convinced “it is always thus.” Even given the evidence of all the phantasmagorical beasties running around, I didn’t believe. So if my assumption-finding process popped out the result that there was something deeply revolutionary afoot, at first I was sure it was wrong. It was only after working out the consequences—essentially all the material in the next three chapters—that I started to believe this formulation was correct. This experience of my own stubbornness, (which I’m deleting from this account for brevity) informs the next comment. ↩︎
