A world without “ESR”

One of my commenters speculated as follows:

Perhaps I overestimate him, but I suspect that without Eric our choice would be Richard Stallman or Bill Gates without much in between. That isn’t a pretty picture. Maybe Linus Torvalds would have help fill the vacuum, or perhaps someone else would have stepped up.

Because I think at least part of the time like a historian/anthropologist, I’ve actually spent a fair amount of effort contemplating what the world might look like if I hadn’t affected it. The more general and interesting question this touches (and what makes this particular instance actually worth thinking about) is a familiar one in historiography: to what extent the times make the man versus the man making the times.

At one extreme you get Thomas Carlyle’s view of history, in which heroic figures shape events and the whole drama is basically their psychologies and struggles writ large. In this version, the details of “ESR” (my public persona) were spectacularly important and a different culture hero would have led to a radically different outcome. At the other you get the Marxist style in which vast tides of history sweep us all along and “heroes” do what they had to do because they had to do it. In this version, I was nothing more than a conduit or focal point for changes that were going to happen anyway.

Of course, I don’t have a certain answer to this question…but in some ways I’ve been a uniquely privileged observer of the historical processes in which I was personally involved, and it has left me with some thoughts and judgments.

As appealing as it would be to be one of Carlyle’s heroes, I have to say that I come down more on the “times make the man” side. Specifically, I’m pretty sure that the hacker community was going to get a generative theory of open source similar to the one I enunciated in The Cathedral and the Bazaar sometime within five to seven years of the mainstreaming of the Internet around 1994. As it was, a couple of people (Richard Gabriel, Larry McVoy) came damn close to it before I did without quite making it to the key insights. Somebody was going to get the rest of the way if I hadn’t; it was time.

The argument that open source constituted a strictly more efficient mode of production than proprietary development that I made in The Magic Cauldron would certainly have followed on that, and a propaganda campaign aimed at selling that truth to people with lots of money would have followed at one remove. In a general sense I think all these developments were pretty much locked in as soon as the Internet began to show people what sorts of productive organization are possible when communications costs approach zero asymptotically.

So, was “ESR” unnecessary? What difference did I make? Or, to put it more precisely…if we imagine a sheaf of histories in which different hackers made that inevitable conceptual breakthrough, what’s marked about the ones in which it was me? In what ways did we (the hacker culture, civilization at large) get lucky….or unlucky?

Most days I think we did in fact get lucky. I think it was inevitable that we were going to get the theorist, and the economist, and the public propagandist. What was contingent and remarkable is that we got one person who could, during a critical transitional period, do all three of those – imagine, and then execute, a change strategy that fused software engineering theory, economic analysis, and effective public advocacy. The result, I think, is that open-source ideas have taken hold rather sooner and rather more strongly than would otherwise have been the case.

I now view the critical transitional period as having spanned roughly 1997-2003. Six years, at the end of which we’d basically booted up a set of sustaining institutions and more or less won the technical and economic argument for open source. My best guess is that without “ESR” we’d have had to grind for a decade, maybe fifteen years, for a fuzzier and more qualified victory.

“But wait…” you might ask. “Isn’t that the hero theory sneaking in the back door, with yourself conveniently cast as hero?” Might sound that way, but here’s where it gets truly interesting. I said most days I think we got lucky. Some days I think something much weirder and more interesting: that there was no real luck involved, that the hacker culture created an ESR when it needed one and because it needed one. This would take us back towards inexorable tides of history and all that.

I am an individual; nobody who’s ever met me is in any doubt about that. But I am also a carrier of culture, my mind shaped by memes. Decades ago, my mind was colonized by a set of memes that you can read about in the Jargon File. Those memes pointed at a kind of perfection – more specifically, at a kind of perfected ubergeeky self that I have been trying to achieve ever since. You could say that I am, at least in part, an invention of those memes. They made me “ESR”…

…and “ESR” happened to become exactly the theorist/economist/propagandist those memes needed to infect others most efficiently in 1997-2003. Wow, what a fortunate coincidence! Or, maybe not. “ESR” may in fact be an especially clever and self-aware memebot, shaped by the hacker culture to propagate the hacker culture. Resistance is useless! You will be assimilated!

Now we have, at least implicitly, two possible theories that bear on the original question. In one, I’m a culture hero; in the other, a sort of memebot/paladin. The practical difference is this: if I was in relevant ways invented by the hacker culture, the hacker culture can invent another paladin like me…and probably would have within a few years if I hadn’t emerged into the role around 1996.

Another way to put this is that if the “times make the man” theory is really strongly true, and the hacker culture invented me, we would not in fact have been stuck with “Richard Stallman or Bill Gates without much in between”. The in-between would have happened, just with a different set of initials.

(It wouldn’t have been Linus Torvalds, or any plausible alternate version of him. He has the brains for the role but the wrong psychology – too much the classic introverted geek, not enough performer/communicator.)

Now I’m going to admit that I fibbed a little earlier. I do not in fact believe one of those theories some days and the other the rest of the time; that was in the nature of a rhetorical flourish. I actually believe them both at the same time. The weighting coefficients in that 2-vector do fluctuate, however. Most of the time, the arrow lands on the historical-determinism side of the x=y line. But not always.

I am both inventor and invented. I mostly think a hacker culture without ESR would have grown something rather like one. But the details and the style would be different. Less Heinlein influence, probably, and fewer guns – though I do think the other-timeline versions would be mostly libertarians, because without starting from a Hayekian stance some key elements of the theory would be rather more difficult to notice. It certainly wasn’t entailed that our paladin had to be a neopagan or have cerebral palsy. On the other hand, I’d bet on at least a large minority of my other-history analogues being shadow-Tourette’s cases with an interest in martial arts; those traits contributed to my toolkit in subtle but important ways.

Sometimes I wonder, though, when the coefficients fluctuate that way. Maybe I really was a unique, heroic axis of history a la Carlyle. It’s an interesting thing to be unsure about.

113 comments

  1. I think that the culture would have produced a persona that would have filled your cultural meta-purpose, but I am equally certain that person would not have brought the unique individual personality to the table which actually makes your general writing (e.g., this blog) interesting to read and thought provoking.

    Your life experiences and upbringing (as I understand it) have certainly gone a long way to shaping you into a very unique individual. Hacker culture is richer for it, but so is American Culture (which I am not sure would have happened if someone else would have been in your “shoes”).

  2. Nice post. Minor nit, however.

    On the other hand, I’d bet on at least a large minority of my other-history analogues being shadow-Tourette’s cases

    I assume you’re talking about the coprolalic symptoms sometimes associated with Tourette’s. Tourette’s isn’t actually what you see portayed in movies and television; it’s a nervous tic disorder. Those symptoms are actually only present in a minority of cases (~10%), and it’s almost never the case that someone with Tourette’s has coprolalia as their sole tic.

    Other than that, I’m not sure I agree with your assessment.

    1. >I assume you’re talking about the coprolalic symptoms sometimes associated with Tourette’s.

      No, I’m not coprolalic – in fact, my use of profanity in speech is fairly light compared to my peers, and almost always for deliberate rhetorical effect. I was thinking more generally of the twitchiness, excitability, and lateral thinking that Robert Sacks says are found in a range of dopamine-excess personality types extending towards Tourette’s at the clinical-pathology end.

      Sacks notes that people on this continuum often become performer/communicator/charismatic types. It’s an unusual way for a hacker to be bent, but it does explain a lot.

  3. While that person almost certainly would not have brought *the* unique individual philosophy to the table, I find it very likely that any alt-hist ESR analogue would have *a* unique individual philosophy. People of that variety always do – it’s at least half of what makes them so interesting. Anyone prone to thinking only on the well-worn trails of philosophy isn’t likely to have enough original insights to make it to a position as a prophet/pitchman for a whole new way of doing business.

  4. The reason I don’t agree with your assessment boils down to testicular fortitude. Seriously. Without you, who else would have had cajones to pitch open source to the IT industry as you did? Who else would have essentially gotten Netscape, for example, to understand “open source your code or die?”

    1. >Without you, who else would have had cajones to pitch open source to the IT industry as you did?

      I really don’t think that kind of cojones – passion and courage about transformative ideas – is at all uncommon among hackers. What’s rare among us as the ability to exteriorize that conviction in a way that doesn’t come off as hyperintellectual or plain weird.

  5. One thing is certain: A world without ESR would have had far fewer self-congratulatory blog postings.

  6. This post invites another question, though – where would we be now without Bill Gates?

    Was there anyone else with the vision, the position, and the talents required to create a near-total control of the PC software market?

    Was Microsoft or an analog inevitable?

    Going back further, one could ask the same question about Tom Watson jr. Was he essential to the rise of IBM to its near-total dominance of computing in the 1970s?

    FORTUNE (IIRC) listed the four greatest businessmen of the 20th centure as Henry Ford, Alfred P. Sloan (who created General Motors), Watson, and Gates.

    The last two were in IT, and had a historic effect. But what was it? That is, where would we be without them?

    1. >This post invites another question, though – where would we be now without Bill Gates?

      Very good question, and I don’t think I know the answer.

  7. I find your comments interesting. Did the times make the hero, or did the hero make the times. The plain fact of the matter is that it is always true that the “times” make the hero, because the times make everything. After all our thoughts, our actions, or existence are basically the outpouring of a very large explosion that happened a long time ago, mixed with various random quantum events that were unpredictable, though certainly not rational.

    However, for us, as humans to cope, we have constructed a whole complex system called “free will” that makes us feel a lot better about ourselves. But it doesn’t ultimately change the truth that the disposition of Open Source was primarily decided 13.7 billion years ago.

    However, if we allow ourselves to live in that comfortable human delusion, the simple fact is that the story of history is that the actions of one person do matter a great deal. History brings them to the plate, but the manner and action in which they behave and the choices they make are of huge consequence, whether for good or for ill.

    Circumstances brought about a Jimmy Carter, but Carterism was brought about by the choices Jimmy Carter made. Circumstances brought about Zhu Yuanzhang, but he shaped the defeat of the Mongols and the founding of Ming Dynasty. Circumstances brought about Robert the Bruce, but it was his ingenuity that won Bannockburn and liberated the Scots for many generations. Circumstances brought about MLK, but he set the agenda and manner of racial transformation in the United States, something that many lesser men have tried ever since hard to undo.

    Circumstances brought about Gates, Page, Brin, Jobs, Wozniak, Dell, and Berners-Lee. Without the right circumstances these men would probably have been largely inconsequential. However, given the circumstances, what they chose to do dramatically altered everything. Different men or women would no doubt have done different things, and the results would have been dramatically different.

    Circumstances provide a stage, but the actors make the ending of the drama.

    Don’t underestimate yourself. (God, I never thought I’d say that to you of all people!)

    1. >Don’t underestimate yourself. (God, I never thought I’d say that to you of all people!)

      That was a funny thing to say, but when I’m wearing my historian hat it’s actually a serious issue. Do I lean towards the historical-determinism theory of “ESR” because otherwise I’d sound like I was boasting of my heroism and I’ve been culturally conditioned not to do that? It’s quite possible. The way of the rationalist is not easy.

  8. Rich, I’ll argue that Thomas J. Watson Jr. was indeed critical to the rise of IBM. Without him, IBM wouldn’t have gotten involved in electronic computing to any sort of degree in the late 40s and early 50s, they wouldn’t have had the balls to bet the business (and that’s exactly what they did) on System/360, and they wouldn’t have ridden the wave as high or as relentlessly as they did. (Don’t forget H. Ross Perot started out a salesman for IBM, and was their highest-profit salesman before he left to found EDS.)

    Where would we be without him? I don’t have a freaking clue.

    1. >Rich, I’ll argue that Thomas J. Watson Jr. was indeed critical to the rise of IBM.

      I have no doubt you’re right, but you may be over-focusing. Forget IBM qua IBM. The interesting question is this: Was it (a) necessarily the case that an effective monopoly would evolve at that early stage of computing, and (b) necessarily the case that the successful monopolizer would be helmed by a personality very like Watson’s?

  9. # esr Says:
    > Do I lean towards the historical-determinism theory of “ESR” because otherwise I’d sound like I was boasting of my heroism and I’ve been culturally conditioned not to do that?

    If it makes you feel any more humble, then your delightful word “memebot” is closer to the real truth, assuming we can get past the whole delusion of free will. However, history is a chaotic system. One tiny, insignificant meme, one genetic mutation, one random[*] event can radically change the outcome. Your cousin Julie telling you that you have cooties when you are five years old could impact the course of world history. There’s a thought.

    [*] by random, I mean real random, not fake random.

  10. > Your cousin Julie telling you that you have cooties when you are five years old could impact the course of world history.

    Finally we know the real reason why Jessica did not write The Cathedral and the Bazaar.

    Yours,
    Tom

  11. Right now, I think I have a front seat to history in the making.

    I am now looking at the dawn of cryptocurrency with the invention of a new money called bitcoin. Bitcoin is a P2P digital currency in which money are issued and tracked by the network as a whole.

    http://bitcoin.org

    The digital currency have many desirable properties of money, such as extreme divisibility, scarcity, difficult to fake and difficult to produce.

    It was invented by Satoshi Nakamoto, who seem to be a cryptography expert from Japan but otherwise not much is known about him.

    In it, I was able to participate as a currency spectator, something I was not able to do. I also participate in auctions and the like. I even brought a share in a new escrow service which will provide 10% of the profit should it take off. Now, I am working on a bitcoin prediction market.

    Did history make the man or the man make history? Will bitcoin even cause a revolution in our economic system? Who knows, but what it is certain is that bitcoin is a very libertarian ideas. It seem that many anarchist hackers have taken up on bitcoin, though there are a few dissenting anarchists who think it is a dumb idea. I called the dissenters and the skeptical libertarians “laughing” anarchists, simply because it remind me of naysayers who thought X won’t work.

    In any case, I bet 500 dollars on the success of bitcoin against a fellow libertarian. 10,000 active bitcoin users in 5 years.

  12. Sorry, got an incomplete thought on the sixth paragraph.

    >In it, I was able to participate as a currency spectator, something I was not able to do.

    It should be…

    In it, I was able to participate as a currency speculator, something I would not be able to do in larger US economy.

  13. > self-congratulatory

    A parallel thread combined with this one has me wondering how many sigmas above mean you have to be before you stop feeling compelled to mock successful people for ever mentioning themselves.

    Blogging about yourself is a natural act. Everybody has self-interest. Pretending otherwise is hypocrisy.

  14. Back in the 80’s, my buddy was a struggling solo music performer. He did OK financially, but really started doing well when he began playing Country & Western music, as the craze hit (remember Urban Cowboy?). He rode that wave for quite a while, until either the wave or his stamina gave out. At the “end” of the “fad”, he was so invested in the genre, that he couldn’t/wouldn’t transition back to rock/folk/whatever he started out as.

    So the times shaped the performer, and the performer(s) shaped the times, and both were shaped by the audience, who were in turn influenced by the fashion of the day and the performers.

    There is no loose end to grab – everything is knit together into a fabric that later on we call history.

    So ESR, I guess the question for you is this:

    If you become a hacker because it was a good paying job, it would be the times making the man.

    If it was because you liked computers/hacking/programming and couldn’t find anything else that you were as good at (as a young man, that is – I’m sure you have acquired other skills along the way), it would be the man making his mark on the times.

    I’m sure there could be other factors, like the hot chick in the math class, or perhaps one truly inspiring teacher, etc…

    1. >I’m sure there could be other factors, like the hot chick in the math class, or perhaps one truly inspiring teacher, etc…

      Dude, when I was a teenager in the 1970s we didn’t have hot chicks in math class. Well, OK, we did, but they were only there under duress. The geekgirl hadn’t been invented yet.

      You geek guys out there who think you have trouble getting dates today….I laugh. I laugh exceedingly. You have no idea what it was like in the pre-Revenge-of-the-Nerds era. You really don’t. And you don’t want to, either. *shudder*

  15. Jessica Boxer> But it doesn’t ultimately change the truth that the disposition of Open Source was primarily decided 13.7 billion years ago.

    Jessica Boxer> …assuming we can get past the whole delusion of free will.

    Why Jessica, I had no idea you believed in manifest destiny :^).

  16. @Morgan – OK… I’ve been in this country for over half of my adult life… and STILL nobody can coherently expla to me What The Heck Are Those Infamous Cooties!?!?!

    1. >What The Heck Are Those Infamous Cooties!?!?!

      Literally speaking, lice. But the term has become detached and mythologized. Good summary here.

  17. If we accept the dichotomy, I tend to favor at least an equal weight to the hero as to the times.

    Certainly the times produce the man — we are undoubtedly products not only of our genetic heritage but our historical/memetic environment. On the other hand, the combination of possessed characteristics – (1) ability to envision a future; (2) ability to grasp the mechanics of reaching it and (3) the ability to tirelessly sell it to the masses surely deserves individual credit — because the synergy is so unique. You need a combination of sufficient intellect + the right temperament + the sheer pitbull determination to make it a reality. How many statistical matches are there for that?

    How different would America have been, e.g. if not for Ben Franklin or Alexander Hamilton?

    But perhaps these are two faces of the same coin, two different ways of looking at the same phenomenon. Maybe the dichotomy is false.

    1. >You need a combination of sufficient intellect + the right temperament + the sheer pitbull determination to make it a reality. How many statistical matches are there for that?

      More than you might think. In the hacker culture, which has a worldwide population on the close order of two million, I think I know of approximately four people with “combination of sufficient intellect + the right temperament + the sheer pitbull determination”. We’re not at all short of people who operate at the required level of intellect, and we’re not really short of people who combine that with pit-bull determination.

      The rare quality in this crowd is the required temperament – we just don’t grow a lot of people who can be described as extroverted charismatics. We don’t even grow a lot of introverts who can fake extroversion well :-). Also, you’re missing a requirement; candidates to be alternate-history ESR have to not have ideological or emotional fixations that would prevent them from being effective pitchmen (e.g., not have Richard Stallman’s problem).

      I judge that those last two qualities (rather than intellect and determination) are the serious statistical filters.

  18. On the same sort of gist, although much much further back, I’ve wondered quite specifically about the rise of Islam; if one probably schizophrenic man had not had a series of initial successes politically, would the entire geopolitical situation even a few centuries later be even remotely similar? If so, it’s probably history’s greatest fluke to what I can tell.

  19. >Without you, who else would have had cajones to pitch open source to the IT industry as you did? Who else would have essentially gotten Netscape, for example, to understand “open source your code or die?”

    In the interests of accuracy, I must note that you’re telescoping some events in a way that may distort causal analysis. Most of “pitching open source to Netscape” required no cojones on my part because I was unaware of the impact my work was having over there. By the time they flew me to California the open-sourcing was pretty much a done deal and what they wanted to talk with me about was execution details.

    You could argue that I exhibited cojones the size of minor planets after Netscape, by consciously embarking on a campaign to convince industry that what Netscape had done was best practice. But as I’ve noted previously in this thread at least twice, I actually don’t consider that sort of testicular fortitude particularly unusual among my hacker peers. We are nothing if not passionate about the Right Thing.

  20. >Blogging about yourself is a natural act.

    This is going to sound odd, but….I think I almost never blog strictly about myself. My style is to use the particulars of my (admittedly remarkable) life as a lens or launchpad to write about questions and ideas that are larger than I am.

    The original post in this thread is a perfect case in point. It only looks like it’s about me; it’s actually about a large question in historiography, using my experience as a probe of the question.

  21. Consider the progress of science and technology. History records the names of those who first publish the results of their research/patent their new designs. There were usually others working on similar ideas, but most people have never heard of them. Perhaps the most striking example was that of Elisha Gray, whose research into telephony paralleled Bell’s. Many historians think Gray was further along, and Bell actually stole Gray’s work. Their patent applications were presented to the government within hours of each other.

    So, if there were no Alexander Graham Bell, we’d still have telephones, but the big monopoly provider would have had a different name. In other cases, without the person who did it first, the innovation could have taken months or years later than it did. It is for this reason that patents are granted only for a limited term (17 years until recently, when it was extended to 20) rather than in perpetuity (as copyrights seem to be heading).

  22. >Blogging about yourself is a natural act.

    This is going to sound odd, but….I think I almost never blog strictly about myself. My style is to use the particulars of my (admittedly remarkable) life as a lens or launchpad to write about questions and ideas that are larger than I am.

    There’s one other aspect of this. One of the truisms about writing is that one should write about what one knows. What better to write about, in that regard, than oneself?

  23. So, if there were no Alexander Graham Bell, we’d still have telephones, but the big monopoly provider would have had a different name.

    And ditto for probably almost all other modern game-changing inventions; the television, electric lighting, the automobile, the airplane, etc. It’s not as if you built a time machine and went back in time and killed Orville and Wilbur Wright right after their opened their first bicycle shop that would cause us to not have discount passenger jet service today.

    Of course, “ESR” didn’t invent open source. He just observed some interesting things about it that most other observers had not. Would others have made the same observations? Probably.

    The question is would others have made the same observations at the right time and under the right circumstances? And would others have promoted those ideas in a way that hit home? Those are answers I’m certainly not qualified to give, and I’m not sure anyone could answer them satisfactorily.

    1. >The question is would others have made the same observations at the right time and under the right circumstances? And would others have promoted those ideas in a way that hit home?

      We also have to beware of hindsight bias, not just in how we answer questions but how we ask them. Notably, your evaluation of a good outcome was in part shaped by me! It is possible in principle that an ESR-alternate with different strengths and values might have led to a very different “right outcome” that, in light of that alternate history, would seem equally satisfactory to alternate you in alternate 2010 in spite of being quite different and unsatisfactory to the version of you in this one.

      It comes back to the heroism vs. tides-of-history question. Did the tribe of hackers follow where the hero ESR led, or was ESR a paladin of hacker memes who simply got out in front of where the culture (the tide of history) was going anyway? I’ve already said that I lean towards the latter view. But my point is that reasoning about this question in terms like right time, right place, and right outcome is very, very suspect – it could be just a way to fool yourself into not noticing that your notion of what’s “right” is one that the hero ESR transvaluated into your head!

      In case it needs repeating, I think that ultra-heroic interpretation is highly unlikely to be true. I’m trying to make a point about the kind of methodological skepticism that is required when we think about such questions.

  24. @esr – “The geekgirl hadn’t been invented yet.”

    I’m sorry but I don’t see how ROTN movie cause the geekgirl phenomenon. Adm. Grace Hopper predates that movie by a significant margin :) And there were geekgirls in other cultures which never had acces to ROTN (i dated one back in USSR – heck, my mom was one way before ROTN came out :)

    1. >I’m sorry but I don’t see how ROTN movie cause the geekgirl phenomenon. Adm. Grace Hopper predates that movie by a significant margin :)

      Sorry, I was unclear. I was proposing ROTN as mainly a marker of change, not a cause of it. It may have reinforced the trend it was marking in a minor way.

      And as for Grace Hopper, it’s certainly true that women with many traits we now think of as “geekgirl” existed before the label. What didn’t, what had to be invented during the 1990s, was the social role of geekgirl – a woman who identifies socially with geeks and announces by labeling herself “geekgirl” that she prefers to hang out with, game with, write code with, and have sex with other geeks. And for such a creature to have enough of a recognized role in the social ecology that other women do not consider her a freak and ostracize her….

      That, trust me, is unprecedented. As recently as the 1970s it was quite unimaginable.

  25. Sorry, for being silly, but after originally seeing the title of this blogpost, this popped into my head and won’t leave until I write it down:

    *Don LaFontaine voice* In a world … without Eric Steven Raymond … hackers and free software are ridiculed and outcast. But one man … will prepare them…. for the fight …of their lives! In 2001, they are placed into internment camps…now the hackers must fight for freedom….and their lives!

    Free as in Freeedom. This movie has not yet been rated.

    1. >Free as in Freeedom. This movie has not yet been rated.

      If I drank coffee, I have no doubt it would be spewing through my nose onto my keyboard right about now.

  26. Morgan Greywolf Says:
    > And ditto for probably almost all other modern game-changing inventions;

    Sure, but who the inventor is, and what he decides to do about it is extremely important.

    Consider for example this scenario: early on Microsoft realizes the importance of Internetworking, and builds it into their earliest operating systems, but makes it entirely proprietary. (I’m not sure the dates work out exactly, but lets assume they got in early enough.)

    So the Internet would have ended as a proprietary invention of Microsoft, with no open standards. Then Linux guys would have spent a lot of time reverse engineering it all much like Samba.

    Internetworking might have been an inevitable development, however, the open, extensible, IPv4 choked Internet we have today certainly isn’t.

  27. We know quite a lot about Admiral Hopper’s accomplishments. Do we know much of anything about her, socially? She has to be unusual simply by virtue of having been a female naval officer…

  28. > We don’t even grow a lot of introverts who can fake extroversion well :-).

    I’ve almost got this down.

    Took me about 20 years to go from hyper-introvert to be able to operate as a mildly smooth extrovert. This was such an unnatural act that I expect I’ve damaged our space-time continuum in some way we have yet to notice the consequences of.

  29. esr Says:
    >>Blogging about yourself is a natural act.
    >This is going to sound odd, but….I think I almost never blog strictly about myself.

    Given that one of the four subjects of this blog listed at the of the page is “software”, and given that open source software is the particular kind of software you are aligned with and interested in, it would be fairly unusual for you to never publicly analyze the main personality who founded the open source movement, whether or not that person was yourself.

    1. >it would be fairly unusual for you to never publicly analyze the main personality who founded the open source movement, whether or not that person was yourself.

      Well put. And touches on one of the most delicate aspects of my situation – that I am both a historian/analyst and a significant actor in the history that interests me. I can’t stop being a historian/analyst, because that’s where my ground as a theorist is. At times I have to distance myself from my own life and my own values in order to maintain the analytical objectivity that allows me to serve those values, and study myself as just another fallible actor in the game!

      If rereading that last sentence makes you feel epistemically dizzy, welcome to my world.

  30. My favorite alternate history is this:

    Microsoft knew, well before the Jackson decision was finalized, that they were facing the possible DoJ mandated dismemberment of the company.

    And it was also pretty well known at the time that Gates had become disgruntled with the entire process of A) running Microsoft B) beating his competitors and C) having to pay tribute to the DoJ.

    I can see an alternate timeline where, in response to a somewhat more hard-assed DoJ anti-trust decision, Gates decides that if he’s going to get torn down by the DoJ, he’ll make the biggest funeral pyre he can imagine.

    First, he sells enough shares that his pile of fuck you money is big enough that he doesn’t need to work ever again.

    Next, he calls RMS and offers him the job of Vice President of Licensing for Microsoft, with a 6 month mandate to make everything in Microsoft fully GPL compliant.

    High probability? Nope.

    Fun to contemplate over beers? Oh yeah…

  31. This post invites another question, though – where would we be now without Bill Gates?

    The result that we have of Microsoft is a result of a few major factors coming together by luck:

    – IBM’s decision to outsource the development of the DOS for the IBM PC.
    – Microsoft’s ability to be supplied by Microsoft (directly or indirectly).
    – The deal Microsoft made whereby which they could resell their DOS to other suppliers as well.

    This resulted in:
    – A separate payment for them for each PC sold.
    – De Facto standardization on their product. When the clones were first made, they were going to need maximum compatibility, and since it didn’t need to be reverse engineered, they resulted in separate sales.

    Resulting in a monopoly position which they’ve managed to hold on to for quite some time.

    Bill Gates is one of the few people with either a consummate business sense or a lot of luck in this area. If he had not managed to push for that deal with IBM, they would still be in the custom software business. Microsoft would probably still be pretty good at it, but they would have missed the big boat. Oracle (the #2 software company in the world) is in the same class as Microsoft (1/2 market cap, but that’s still huge), and it’s possible that Microsoft would have made it. But both of these companies show that it is important to have a single specialty and focus.

    If IBM had decided to in-source the OS, it would have ended up much like the private versions of Unix. It would have made clones harder (at least initially) for clones to be created, but they would have lacked the standardization which ended up as a result of it. It might have been higher quality with IBM focusing more on robustness rather than cost-based market-driven features that Microsoft did.

  32. > In it, I was able to participate as a currency speculator, something I would not be able to do in larger US economy.

    False.

    There are plenty of ways to be a currency speculator in the US economy. Call Schwab or any of their competitors for a small sampling.

  33. Garrett> The result that we have of Microsoft is a result of a few major factors coming together by luck…

    This is true of virtually any successful company or individual (ESR included). “Luck” (right place, right time, right skill set) plays possibly the largest part in all such successes. The trick is, putting yourself in a position to have the three vectors above pointing in your direction.

    The thing I’ve noticed about the self-made people I know is that they ALL work VERY HARD to make themselves the biggest possible target for these three vectors. Most people work on just two of these dimensions but what puts REALLY successful folks into that outlier region is they find a way to work on all three together. Mostly, I think, this is related to the general engineering rule, “fast, cheap, good: pick two,” where there are distinct exception that stand out head and shoulders above the rest (I’ve always thought the P38 can opener was one such).

    WRT Gates vs. IBM, he was in the right place/right time (his mom knew some of the IBM execs that were leading the search for an OS IIRC and sent them Bill’s way), and he had the right skill set. Not only was he a top-notch hacker and businessman, but he had the audacity to push for his retaining ownership of the OS – to reduce the cost of the system to IBM – and the knowledge of what that meant because of his arguments with RMS and crew back in Cambridge about retaining and controlling rights to the code he’d written.

    The difference that I’m attempting to express here regarding skill set isn’t just technical skills, but also psychological skills, the ability to communicate well, and so forth that you don’t get in school (beyond a certain basic point) but have to work at over an extended period of time to acquire. ESR’s upbringing and eclectic curiosity contributes to his ability to do the things we’re talking about here by the bucket full, and his solid chops as a hacker and his ensconced location in the Tribe prior to 1997 or so gave him the place and time to have the opportunity to put that lifetime of odd skills to work.

    Eric estimates that out of 2 Million or so members of the Tribe, there are perhaps 4 with similar skill sets, and that sounds about right to me as well, but one thing that might be lacking in the other three is the willingness not only do a CatB, but to follow it through with “Take My Job, Please” and similar works which disengaged him from the movement before it became a cult of personality. I’d say that would be more like a 50/50 chance, leaving the number of potential ESRs at or near 2.

    In short, Eric, you’re literally one in a million ;^).

    1. >Eric estimates that out of 2 Million or so members of the Tribe, there are perhaps 4 with similar skill sets

      Nit: That was intended as a lower bound – there might be others I don’t know of. However, your general point is correct. We do not have 40 such people; if we did, the ripple effects of their presence would be quite visible.

  34. @esr said:
    If rereading that last sentence makes you feel epistemically dizzy, welcome to my world.

    It’s almost as if you’ve just finished reading I Am A Strange Loop. But I think that’s what Jay Maynard was pointing out. ;)

  35. @dgreer:

    WRT Gates vs. IBM, he was in the right place/right time (his mom knew some of the IBM execs that were leading the search for an OS IIRC and sent them Bill’s way), and he had the right skill set

    I seem to remember it was that IBM was looking for languages, not an OS. When it came down to an OS, Bill Gates sent them to Digital Research, but Gary Kildall’s wife refused to sign the NDAs without Kildall’s approval, so they went back to Microsoft to ask them to do one. Gates said “sure we can do one!” and then went out looking for one and found something called QDOS (Quick and Dirty Operating System) from Tim Paterson at Seattle Computer Products, they purchased it for some obscenely small amount of money and the rest is history.

    However, the rest is spot-on, IMHO. Well put.

  36. >Nit: That was intended as a lower bound – there might be others I don’t know of. However, your general point is correct. We do not have 40 such people; if we did, the ripple effects of their presence would be quite visible.

    It seem that more and more new hackers come with built in economic analytical software. Some even now come with an entrepreneurial approach.(YCombinator)

  37. ESR> Nit: That was intended as a lower bound – there might be others I don’t know of. However, your general point is correct. We do not have 40 such people; if we did, the ripple effects of their presence would be quite visible.

    Yeah, yeah, but then I couldn’t have used my one-in-a-million line ;^).

  38. @ dgreer:
    > WRT Gates vs. IBM, he was in the right place/right time (his mom knew some of the IBM execs that were leading the search for an
    > OS IIRC and sent them Bill’s way), and he had the right skill set. Not only was he a top-notch hacker and businessman, but he had
    > the audacity to push for his retaining ownership of the OS – to reduce the cost of the system to IBM – and the knowledge of what
    > that meant because of his arguments with RMS and crew back in Cambridge about retaining and controlling rights to the code he’d
    > written.

    I completely agree. Please don’t mistake me for one of the “you’re only rich because you’re lucky people”, though I can see how from my previous post you might have come to that conclusion. I tried to make that clear in a later paragraph, but apparently failed.

    The best description for really big success is an intersection of knowledge/ability, hard work, and luck.

    There are some opportunities that anybody could have latched on to. The iPod phenomenon – Apple wasn’t the first, but they were the “best” as seen by a large percentage of the population. That wasn’t luck, that was a large number of sets of abilities combined with dedication and drive. Same with Facebook – lots of people tried before, but Facebook has been the most successful due to hard work and getting it right (yes, I’m ignoring the controversies).

    What made Microsoft was definitely knowledge and hard work. Everything I’ve read indicates that Bill gates is both a first-class technical wizard as well as a first-class businessman. But if he’d come along a few years later and missed the opportunity with IBM, or if Gary Kildall had been home at the time and missed the provider+licensing model that he went with for DOS, Microsoft wouldn’t have been the company that it became. Maybe it would have owned the embedded space much like VxWorks did for years, maybe it would have been nonstop killer office software. I have no doubt Microsoft would have done well, but it would not become the company that it was. If big opportunities like that were readily available, there would be far more founder-led software companies in the 100 billion dollar market cap range. There aren’t.

  39. One thing I’ve been thinking about since my last post is how much all of this reminds me of similar success stories throughout history – Carnegie, Rockefeller, Westinghouse, and Franklin come immediately to mind – and it makes me wonder: how could you possibly recognize these sorts of opportunities as they appear?

    I doubt you’d be able to see them before they appear because they seem to be sort of a perfect storm of events happening at a given time and place.

    Of course, I think maybe ESR’s perfect storm may have stuck around for a while. From what I recall, RMS tried to step into that space with FSF, but he didn’t quite hit the nail on the head (but he pointed in the right direction), so I think the pot may have kept boiling until the RIGHT hero showed up with the RIGHT idea to finally give the Tribe the direction it needed.

    Oh well, there goes my analogy :^).

    1. >Of course, I think maybe ESR’s perfect storm may have stuck around for a while. From what I recall, RMS tried to step into that space with FSF, but he didn’t quite hit the nail on the head (but he pointed in the right direction)

      I learned a great deal from RMS’s example. Unfortunately, much of what I learned was of the form “don’t do that!”

  40. > I was thinking more generally of the twitchiness, excitability, and lateral thinking that Robert Sacks says are found in a range of dopamine-excess personality types extending towards Tourette’s at the clinical-pathology end.

    I think you mean Oliver Sacks.

    1. >I think you mean Oliver Sacks.

      You’re correct. I had a collision in my hash tables with a Robert Sacks I used to know.

  41. “The rare quality in this crowd is the required temperament – we just don’t grow a lot of people who can be described as extroverted charismatics. ”

    I’m having trouble figuring out your MBTI type, something that doesn’t happen often.

    My guess is it’s either “Promoter/Persuader” (ESTP) or “Inventor/Originator” (ENTP).
    The charismatics are often ESTPs, but you have quite a bit of rational NT emphasis.
    Maybe the S/N function is borderline and you shift between both.

    “Also, you’re missing a requirement; candidates to be alternate-history ESR have to not have ideological or emotional fixations that would prevent them from being effective pitchmen (e.g., not have Richard Stallman’s problem).”

    I’ve noticed that – there’s a certain openness that such a candidate possesses which engenders trust.

    Perhaps the weighing coefficients you mention reflect the relative proportion of “struggles” vs. “opportunity” (to utilize natural gifts – which implies “times make the men” or “give him a playing field”.) Might they not be different for different historical figures?

    1. >I’m having trouble figuring out your MBTI type, something that doesn’t happen often.

      ENTP. Strongly ENT, weakly P.

  42. esr>Unfortunately, much of what I learned was of the form “don’t do that!”

    I’m really glad you learned the lesson well :^).

    1. >I wonder if any of these people read The Cathedral and the Bazaar.

      Yes, but it only reminded them of what they’re supposed to be doing as scientists anyway.

      I’m not being flip by saying that. I know the following things to be true:

      1. When I began to understand the importance of source transparency and many-eyeballs skeptical review in software, it didn’t take me long to recognize that the open-source self-correction process is closely analogous to (if not identical to) scientific peer review. I frequently presented it in those terms.

      2. For reasons largely having to do with the desire for revenue capture from IP monopolies, since 1980 scientific researchers have been withholding holding more and more of their data from open publication and scientific publishers have been clamping down on the redistribution of scientific papers.

      3. There has arisen in the last decade an insurgent open-science and open-publications movement pushing back against this trend. Physics and bioinformatics are two fields where the insurgents have been especially vocal and active.

      4. Where you can find manifestos from the open-science insurgents, they do have a strong tendency to refer to the example of open-source software and cite my work.

      Need it be said that I find this rather ironic? I’m only reminding scientists of their own best practice. But, hey, I’m glad I’ve helped supply the insurgents with ammunition, even if I’m really just handing them back their own bullets.

  43. @Ken Burnside
    > Next, he calls RMS and offers him the job of Vice President of Licensing for Microsoft, with a 6 month mandate to make everything in Microsoft fully GPL compliant.

    …but he gets distracted trying to change the name of the company to GNU/Micorosft and nothing ever comes of it!

    @Jessica Boxer
    > Consider for example this scenario: early on Microsoft realizes the importance of Internetworking, and builds it into their earliest operating systems, but makes it entirely proprietary.

    That’s actually how networking was done in the early days. Many companies had their own proprietary networking systems. DEC with DECnet. Xerox with XNS. Apple with AppleTalk. Unix with UUCP. IBM with SNA. Novell with Netware.

    @Morgan GreyWolf
    > […]so they went back to Microsoft to ask them to do one. Gates said “sure we can do one!” and then went out looking for one and found something[…]

    Some biographers/historians who have written of this say that Gates told IBM that Microsoft didn’t have an OS, and that one of the top guys from the IBM side took them aside and said “if you don’t propose supplying an OS, this project is going to be cancelled”. At that point, Gates has nothing to lose, so says they will supply an operating system.

  44. One of the things that always attracted me to OSS in the first place was its decentralized nature. Linus, ESR & RMS could expire in a plane crash and GNU/Linux would soldier on unimpeded. Some may argue even less impeded, but I digress!

    Historically, there is a pretty clear case that not much would be different in a world without ESR. As per my favorite open-source project WikiPedia, there is a rich history of knowledge sharing that predates the information age:

    http://en.wikipedia.org/wiki/Open_source#History

    Unix itself was open-source from the beginning. BSD itself was derived from source code donated by Ken Thompson to UCB while on sabbatical there in the 70’s. ESR was merely a historian documenting a well established phenomena.

    In response to ESR’s numbered comments above:

    1. The scientific peer review process is demonstrably superior to that of open source, as the barrier to entry is much higher. When you read a paper in a reputable, peer-reviewed journal you can be reasonable certain that the author and reviewers are competent professionals. The open source process makes no such guarantee and as such its quality varies greatly; given the attention paid by incompetent amateurs.

    2. There is one golden rule in the marketplace; he with the gold makes the rules! Investors that sponsor research have the right to decide what to do with it. Researchers have the right to decide whom they want to work for. We as consumers have the right to decide what we want to support with our purchasing dollars. This is the free market at work.

    I’m even a perfect example of this. I chose to work for a large, public research institution as I have more freedom to share information with my peers. I pay for this freedom by accepting a lower salary than I would working within a proprietary system. This is of course fair and simple market economics.

    3. This is simply not true. The National Science Foundation was created in 1950 and public research has been present throughout history. The libraries are full of ‘open’ research. And if by ‘open’ one means subject to the same low standards which are symptomatic of open-source software, this is by no means a good thing.

    4. Thankfully, it appears the ‘open’ review process is inheriting the best features of both open source and classic professional peer-review. The barrier to entry is the same, with simply a more transparent process. This by any means is a good thing.

    In closing, I’ll add that the modern open-source source model has really produced little in the way of technical innovation vs. proprietary software; so to that end I think its history is still being written. Stop copying and start creating, folks!

  45. @Ken Burnside
    > Next, he calls RMS and offers him the job of Vice President of Licensing for Microsoft, with a 6 month mandate to make everything in Microsoft fully GPL compliant.

    …but he gets distracted trying to change the name of the company to GNU/Micorosft and nothing ever comes of it!

    I don’t normally do this…but I can’t resist:

    Thread over.

  46. It seems to me that by 1997 the ‘open source’ movement was fully underway and that “ESR” acted as a sort of Internet-enabled town crier who spread the news, but did little of consequence from a software or business perspective. For example, many things predate 1997: GNU Project (1983), Perl (1987), Linux (1991), FreeBSD (1993), Apache (1995). And even open source corporation pre-date 1997: Cygnus (1989), Red Hat (1993). And the roots of open source lie even deeper than that.

    My take is that “ESR” was unnecessary and that open source would have happened without that role. By the time CatB came out the entire movement was on a roll.

    1. >My take is that “ESR” was unnecessary and that open source would have happened without that role. By the time CatB came out the entire movement was on a roll.

      One of the ironies of a victory as complete as I actually scored was that it can make the very fact that any struggle was necessary invisible in retrospect. But that’s OK; it’s actually the outcome I was hoping and planning for, and occasionally predicted would occur.

  47. John, your analysis leaves out two salient points:

    1) “Open source”, as a concept, didn’t exist as such. What there was was so-called “free software”.
    2) The only spokesman any of this had, really, was RMS.

    Without the role Eric filled, the latter would have held back the whole idea, perhaps fatally, because it would have been associated with RMS’s kookiness. RMS is incapable of seeing that his zealotry and insistence on people not only doing the right things, but doing it for the right reasons, harms his cause far more than it helps. He’s surrounded by simon-pure true believers, and doesn’t see that the shell is a lot thinner than he thinks.

    1. >Without the role Eric filled, the latter would have held back the whole idea, perhaps fatally,

      Not fatally. I stand by my previous assessment that my absence would have cost us roughly four to ten years. Of course, that’s enough time for an awful lot of proprietary mischief….

  48. Not fatally. I stand by my previous assessment that my absence would have cost us roughly four to ten years.

    I agree. I simply see that delay as being at least highly damaging, f not fatal, by giving other solutions a chance to become entrenched. How long did it take Windows to lock down the marketplace?

    1. >I agree. I simply see that delay as being at least highly damaging, f not fatal, by giving other solutions a chance to become entrenched. How long did it take Windows to lock down the marketplace?

      Damaging, yes, but not fatal because in the long term the economics was on our side. You know this, but like many hackers you have some tendency to get so caught up in the virtuous struggle against this year’s bad guy that you lose sight of the longer-term positive trends.

  49. > John, your analysis leaves out two salient points: 1) “Open source”, as a concept, didn’t exist as such. What there was was so-called “free software”. 2) The only spokesman any of this had, really, was RMS.

    I agree with that, although I agree most strongly with ‘free software being a bad concept primarily because the word ‘free’ has specific connotations and was being used in way that most people would not understand (hence the need to repeat free as in speech, not as in beer all the time). Hence “Free Software” was bad marketing, whereas “Open Source” was much better.

    But I still don’t believe that “ESR” was necessary. In fact, my claim is that he was unnecessary and that the groundswell of ‘open source’ was so large and growing that it was bound to occur at or around the time it did.

    The other problem with the claim is that “ESR” didn’t invent the term ‘open source’, not even the ‘open source definition’ (which is essentially the Debian Guidelines). So, I have a hard time seeing what he did do that was of significance other than act as the ‘town crier’ who spread the news.

    > One of the ironies of a victory as complete as I actually scored was that it can make the very fact that any struggle was necessary invisible in retrospect

    So what were the struggles?

    1. >So what were the struggles?

      I could list many, but for now I’ll just cite the single largest one: convincing corporate kingpins that open-sourcing code made sense as a sustaining business strategy because it increased the efficiency of production. Before we got to that point, I had to disrupt two other positions: (1) Open-sourcing is just crazy, you’re giving away the IP crown jewels, and (2) open-sourcing only makes sense as a move to disrupt competitors. While it would not become clear to everybody else for about three years afterwards, I knew we had turned the corner on this one when I successfully co-opted IBM’s strategic planners in late 1998/early 1999.

      There was a subtler flip side to this one. In order to get the hackers to a point where they would actually hold up their end of the new game, I had to sell cooperation with markets and capitalism to a culture that had a long history of reflexive suspicion about them. This was not as difficult, but it took some effort and careful reframing. I had to alter the discourse of hacking, the set of implicit categories within which my peers formulated their choices, away from one that promoted fear of corporate predators and a reflexive oppositionalist crouch (as in, for example, the Free Software Foundation’s propaganda) and towards one in which hackers saw themselves as a community with enough leverage to trade as equals in the market.

      You implicitly criticized me for having picked up the DFSG and used it to define a gold standard for open-source licensing, but you’re missing a very large point. When you’re trying to social-engineer the direction of an entire culture, you have to use the cultural capital that’s available – if you try to get too far out in front of the culture, what you do will come across as disconnected radicalism and you’ll fail. Thus, the cleverest transvaluations always look trivial or inevitable in retrospect. There’s a famous line from chapter 17 of the Tao Te Ching that expresses this: ‘When the leaders complete their work, the people say, “We did it ourselves.”‘

      I actually had to re-engineer two cultures at once – hackerdom and the corporate world. It is emblematic of the degree of success I achieved that in 2001, the book version of CatB (a book about the behavior of hackers and geeks!) hung out on the New York Times business-books bestseller list for a while. We can take that sort of positive mainstream exposure for granted today, but only because I broke the ground.

      Which is not to say these things wouldn’t have happened without me, eventually; they would have, I’m convinced of that. But I it would have been a much slower and messier process with more wastage and screaming and foot-dragging on both sides. It’s good that we were all spared that.

    2. >So, I have a hard time seeing what he did do that was of significance other than act as the ‘town crier’ who spread the news.

      Replying on a different level than my previous: did you miss, or manage to forget, what I did to the foundational assumptions of the theory of large-scale software engineering in 1997? Don’t take my word for this; read what Fred Brooks, who poured the concrete for those foundations in the 1970s, has to say about my work in The Art of Design. Later, I wrote a novel generative theory of the economics of software that has spawned an academic cottage industry of followup papers, not to mention every open-source business model in existence.

      Before Brooks and myself there was David Parnas’s work on modularization. Now try and come up with anyone besides Parnas, Brooks, and me who is described by his peers as a major theorist of scaling in large software systems. Next, name anyone besides me who has written a generative theory of software production economics that others could turn into actual business plans. I mean, one which gets down to the predictive level of explaining which pricing models do and don’t work.

      “Town crier.” Hmmph. I don’t mind having my leadership forgotten; I aimed to make myself inessential on that level. It does nettle me a bit when people forget that I’ve been a pretty effective theoretician, too. I suppose I shouldn’t complain, though; I knew this was an ancillary risk of becoming a public pitchman.

  50. >So what were the struggles?

    I’d have to suggest that convincing business school types that the long-term payoff of open sourcing a project was worth the short-term downside, and over coming the public perception (and FUD) that if people could see the source code the software would be less safe. I know that’s been the two main points of problems from my own experience, although I’m pretty sure that Eric had to deal with this and more.

    He became a leader of the movement and as such had to try to wrangle, cajole, and shove everyone in more-or-less the same direction to get things accomplished. THAT’S the job I wouldn’t want. I can evangelize about OSS all day long (and have spent much of my career doing that), but when you’ve got the FS vs OSS religious battles and similar harebrained things going on, it creates a lot of little fires that have to be put out, and having his face on one of those two totems made Eric a significant target from WITHIN the FOSS community, never mind the other side’s arguments.

    Yes, we would have gotten there eventually, through natural selection if nothing else, but how much better off are we that it happened in 10 years rather than 20? And what great products do we have today that likely either wouldn’t exist or would be a WHOLE LOT less affordable? Can you remember when “Soho” routers were $1000? Can you remember when a wireless AP cost $800-1200? Can you imaging the little picture frames out there being $40-50 if they each had to develop or license a proprietary OS (I would direct you to the GPS market that resisted Linux for proprietary code until about 5 years ago and maintained a cost of $300+)? Do you remember when a Gbps switch was $100K and pretty much Cisco and Foundry were the only games in town?

    These things happened because we got the manufacturers to accept Linux and BSD as acceptable platforms for their equipment, because we got companies like IBM to help improve the product (linux, etc.), and release some of their patents and in-house code to better the product for everyone. THAT was the struggle. Netscape was a very visible win, but it was only one win, and ultimately I don’t even think it was the biggest one.

    In the networking world, I see the results of the advance of OSS as a platform on a daily basis. I’ve got a switch sitting in my DC that lists around $150K, but if it was a similar Cisco it would be twice that (maybe 3 times), and it out performs the equivalent Cisco on most things, including raw switching speed. Virtually the entire wireless industry is now based on Linux, as is most of the VoIP industry. High-end routing, outside IOS, runs almost exclusively on custom build BSD platforms. Huge companies have made the switch OSS based platforms (IBM, HP, Motorola, Samsung, Apple, NSA, DOD) and contributed back to the community man millennia of development work .

    We are significantly better off today because the change happened when it did, and much of the credit for doing the heavy lifting goes to Eric and the team he lead. RMS couldn’t have done that, because he never would have gone to IBM or Netscape and sat down with them to discuss the topic in even tones, discussed who it helped their bottom line, and that’s true of most folks who evangelized for FS prior to OSF’s founding – they tended to come off as a bunch of zealot communists. Eric made the argument one of free market dynamics, put it into business school terms, made it understandable by “the suits”. That’s not an easy thing for those with the hacker bent.

    1. >Netscape was a very visible win, but it was only one win, and ultimately I don’t even think it was the biggest one.

      I agree. It was the shot heard ’round the world, and it had the useful tactical effect of making my humble-or-not-so-humble self credible in Silicon Valley and on Wall Street, but the important wins were the ones in which the hackers collected on that advantage.

      Of these, I still think bringing IBM over to our side was the single most important. Oracle’s slightly earlier pioneering decision to make Linux a tier-1 platform was pretty important as well. I was directly involved in the former but not in the latter.

  51. I agree most strongly with ‘free software being a bad concept primarily because the word ‘free’ has specific connotations and was being used in way that most people would not understand (hence the need to repeat free as in speech, not as in beer all the time).

    Free clue: If you have to explain what your core word means in order to get people to apply your meaning instead of another one, you’re engaging in argument by redefinition. It’s not just bad marketing. It’s fundamentally dishonest, and it’s wrong.

  52. So, I guess we get to a fundamental issue. You say that you were responsible for changing the corporate world and hackerdom. Who else would? (of any importance). If we asked people from OSI, or IBM, or any other company you name, would they believe that?

    I agree that CatB was an important book, but that fact doesn’t seem to bolster your importance because it can also bolster my argument that you were the ‘town crier’ who told the story.

    If you read IBM’s official history of their involvement with OSS (A history of IBM’s open-source involvement and strategy published in IBM Systems Journal) you get mentioned because of CatB (alongside “Open Sources” from O’Reilly), but IBM states that they were seeing Linux adopted by their customers and that that’s what drove them to truly understand open source and how it applied to their strategy.

    1. >You say that you were responsible for changing the corporate world and hackerdom. Who else would? (of any importance). If we asked people from OSI, or IBM, or any other company you name, would they believe that?

      Well, that’s an excellent question for you to be asking other people, isn’t it?

      I mean, I could do an extended song and dance about this to try to convince you, and I could in fact name a lot of witnesses, but from my point of view there wouldn’t be a lot of gain in doing that. I wasn’t after adulation, I was after victory. Whether I got the credit for it afterwards didn’t figure into my thinking much then, and still doesn’t today. So you go right on thinking of me as a mere “town crier” on that level if you like; it won’t change anything that I accomplished one bit. You’ll still be living inside a discourse that I did much to create, and I’ll know that even if you don’t.

      I do wish you wouldn’t forget that I wrote some important theory, though. That I do want remembered.

  53. esr> I knew this was an ancillary risk of becoming a public pitchman.

    I hear their looking for somebody to hock Sham Wow. You could be the next Billy Mayes. ;^)

  54. And here’s where I display my ignorance. I don’t know much about the theories you are talking about and ought to educate myself. What’s the best thing for me to read on the subject?

    1. >I don’t know much about the theories you are talking about and ought to educate myself. What’s the best thing for me to read on the subject?

      The software-engineering side is relatively easy. Start with David Parnas’s foundational paper On the Criteria To Be Used in Decomposing Systems into Modules; you may find it a little difficult to take seriously because so much of it has now passed into conventional wisdom, but the point of this exercise is to put your mind in place where what he was writing about could seem novel and groundbreaking. It really was in 1972.

      Then read Fred Brooks’s The Mythical Man-Month, which became the most important single essay on large-scale software engineering. The book is full of insights, but the most important one was the quadratic scaling law of software project communications overhead. What I did, on the theoretical level, is best understood as a reply to Brooks’s original assertion of a quadratic scaling law. I showed that this depended on assuming that projects need to pay the O(n**2) complexity overhead on the complete graph of n communicating programmers, but that a starlike communications structure could reduce that cost function to being asymptotically close to linear. The key point is that organizing this way inverts Brook’s prescription – adding more programmers yields a net gain instead of swamping you in overhead.

      The economics side is more difficult, because you have to understand how to do what Austrian economists call “praxeological analysis” first. I’d recommend you start with David D. Friedman’s Hidden Order: The Economics of Everyday Life first; if you’re feeling brave, get the version of that book called Price Theory that leaves in the math. David’s basically a neoclassical economist rather than an Austrian, but if you pay attention his book will show you how to do praxeology.

      My work on open-source business models is applied praxeological analysis starting from the incentives on all parties created by different pricing models and different degrees of information asymmetry. There isn’t anything else quite like it in the economic literature, because since about 1950 most economists have had a bad case of physics envy that they expressed by turning everything into calculus about various forms of (mostly fictional) equilibrium. The closest modern parallel to my work inside the academy is David Friedman’s own stuff on the economic foundations of legal systems. This is not coincidence, David and I are friends and share a lot of intellectual ancestors.

  55. The key point is that organizing this way inverts Brook’s prescription – adding more programmers yields a net gain instead of swamping you in overhead.

    What does Brooks think of this idea, and how come he didn’t try it?

    David and I are friends

    And you let him release Harald anyway? I got it at the dollar store, and…meh.

    1. >What does Brooks think of this idea, and how come he didn’t try it?

      He has applied the adjective “brilliant”. As for why he didn’t try it, he did try scaling up and observed horrible results. What he didn’t try was radically decentralizing a development the size of OS/360 or the (rather larger) Linus kernel, collecting the advantages of a starlike communications structure. In 1975 when he wrote his book, the technology to support this was only beginning to become imaginable; it hadn’t even been that when OS/360 was being planned in the early 1960s.

  56. I showed that this depended on assuming that projects need to pay the O(n**2) complexity overhead on the complete graph of n communicating programmers, but that a starlike communications structure could reduce that cost function to being asymptotically close to linear.

    Couldn’t the O(n^2) organization be called a community of equal designers, whereas in a starlike organization, the single hub in the center represents a higher authority?

    In other words, committees suck at design.

    1. >In other words, committees suck at design.

      Yes, but the more positive spin is while design doesn’t parallelize well, bug characterization (and to some extent, exploration of the design space) actually does.

  57. Hm. What technology do you see as central to enabling the decentralized development structure? I’ll agree that it almost certainly wasn’t available in the 1963-1964 timeframe, when OS/360 could have made the best use of it; I’m just curious if it was even possible on the systems of the time, or if equivalents could have been developed. There’s a difference between imaginable and possible, I agree; I’m thinking in terms of possible.

    As for the size of OS/360 vs. the size of the Linux kernel…hrm. More people work on Linux, certainly, but I should run a wc -l on the OS/360 source code to see how big it is.

    1. >What technology do you see as central to enabling the decentralized development structure?

      Personal workstations powerful enough to run a full development environment without stepping on anyone else’s instance and cheap, fast networking to connect them. Today, you can hack a full-blown operating system with equipment that sits on your desk. This wasn’t really possible before the mid-1980s, though the earliest experiments with such workstations date to about ’73.

      >I should run a wc -l on the OS/360 source code to see how big it is.

      You should. I’d like to know that statistic.

  58. >What technology do you see as central to enabling the decentralized development structure?

    Personal workstations powerful enough to run a full development environment without stepping on anyone else’s instance and cheap, fast networking to connect them. Today, you can hack a full-blown operating system with equipment that sits on your desk. This wasn’t really possible before the mid-1980s, though the earliest experiments with such workstations date to about ‘73.

    Hm. You can get there from here as early as CP/67, plus or minus the inter-system networking. You can run multiple instances of CP/67 under CP/67 easily; it was designed specifically to allow that. The gotcha there is that you can only run some number of them on one host, and BITnet (which, don’t forget, started life on VM/370 systems) may or may not be powerful enough.

    >I should run a wc -l on the OS/360 source code to see how big it is.

    You should. I’d like to know that statistic.

    The complete OS/360 version 21.8 distribution, the only one I have access to (and possibly the only one still extant), clocks in at 3649361 lines of code, including everything that came with it (utilities, compilers, and so on). It’s all tolerably well commented assembler.

    1. >BITnet (which, don’t forget, started life on VM/370 systems) may or may not be powerful enough.

      I used BITnet briefly. It wasn’t. :-)

      >The complete OS/360 version 21.8 distribution, the only one I have access to (and possibly the only one still extant), clocks in at 3649361 lines of code, including everything that came with it (utilities, compilers, and so on).

      So the whole OS/360, userland and all, was about a quarter the line count of a current Linux kernel. Interesting.

  59. How much of that 15 million or so lines of Linux kernel is device drivers for damned near everything that could be hooked up to a computer, and how much of that is what the average person would think of as the OS kernel?

  60. OK, let’s start with “one in a million” and count down by orders-of-magnitude from six billion; 600 million, no lasting effect; 60 million, . . . and at one-in-million for esr, we’re at 6,000 of people alive now. I do not know Eric personally, but I knew his books (CatB) before discovering the blog, and 6,000 of the world seems pretty reasonable. I don’t know how you’d narrow it down between 600 and 60,000.

    What I find absolutely fascinating about this thread, and the earlier one about “the curse of the gifted” is that people are calmly and rationally discussing who could have changed the world and how smart they had to be to do it, and not getting all hung up in discussions about whether it is socially appropriate to discuss intellectual ability at all.

    I don’t want to hijack Eric’s comments, but I’d welcome email (on web page under “about”).

  61. Linda Seebach: The “one-in-a-million” was more of a joke than anything, but if you want to do an analysis, you start from an assumption that everybody in the world is qualified to have done what Eric did, which is untrue. From the outset, the discussion of potentially qualified individuals was restricted to Hackers and would-be Hackers because that is the specific sub-domain of individuals who had the visceral knowledge of WHY OSS matters.

    As for intelligence, in order to do the sort of formative thinking required for the things that happened that are discussed in this thread, there’s a basic barrier to entrance in effect. It might be possible to get around part of this barrier with specific knowledge (e.g. somebody who studies economics could understand the economic repercussions of the topic) but to have to cross-discipline knowledge to a depth and breadth necessary to see all the pieces in CatB before CatB would require a genuine polymath, and for that there’s a definite minimum level of intelligence in effect.

    Why would it be socially unacceptable to have such frank discussions on a forum that is, by and large, populated by people to whom such things are of particular and personal concern? Would you expect a blog dealing with sex addiction to not have frank discussions about sex needs and gratification?

  62. >CatB would require a genuine polymath, and for that there’s a definite minimum level of intelligence in effect.

    My observations of the geniuses I’ve hung out with and read about suggest that you start getting serious polymathy at about IQ 150. Admittedly that’s on a small sample, but that far off the mean there are no large samples.

    On the other hand, I don’t think I was operating at the limit of my ability when I wrote CatB. The breakthroughs in that paper had more to do with breadth and cross-disciplinary perspective than heavy thinking; I remember that once I decided to write it, the effort involved did not seem very large. And what effort there was mostly involved summoning up the willingness to write down the startling conclusions of my logic.

    The reason I’m discussing this is that we can put it together with a conjecture from a previous thread. While I don’t have a precisely measured IQ, never having been tested with an instrument calibrated for the right range, the most plausible estimates seem to cluster in 160-165.

    Therefore: I’d express the minimum IQ required for an alternate-history ESR as 155 plus or minus 5. I’ve lost my pointer to the the frequency table I was using in the Curse of the Gifted thread, or I’d translate that into a population-incidence range. Does anyone listening remember enough applied statistics to do that?

    It would be interesting to see if it overlaps Linda Seebach’s estimate of 6,000 candidates on the planet, which I agree seems plausible.

  63. esr> The breakthroughs in that paper had more to do with breadth and cross-disciplinary perspective than heavy thinking…

    That’s probably because you’d spent a significant portion of your life doing the analysis up to that point, in bits and pieces, and never sat down to put it all together. I’m certain, with what I know of you, that you didn’t fall into the “Free is Good” camp without introspection (and you don’t strike me as the person that goes there because it’s “cheaper” :^), and I’m equally certain that you didn’t read about economics or study business and movement patters without an eye toward how that knowledge applied to what you already knew (the free software/hacker movement), so the lack of mental effort was because you’d already thought most of the hard thoughts. You simply pulled those thoughts together and figured out what conclusions they lead to. CatB wasn’t the result of a couple of weeks work, it was the result of years of work, and that’s why you’re “one-in-a-million”.

    1. >That’s probably because you’d spent a significant portion of your life doing the analysis up to that point, in bits and pieces, and never sat down to put it all together.

      It would be correct to say that I spent a significant portion of my life assembling the knowledge base and mental toolkit that I deployed in that work. But it’s not obvious that this fact alone consitutes evidence of one-in-a-million specialness in intellectual ability. After all, isn’t this sort of mental preparation what scientists and scholars are supposed to do with their lives? And we certainly produce those at a greater frequency than you are supposing we produce ESR-equivalents.

      I mean, it’s nice that some people think that well of me, but I don’t think it’s a realistic assessment in comparison with theorists in other fields. Compared to (say) the Weinberg/Salam formulation of the Standard Model of physics, or the recent proof of the Poincare theorem, or (if it’s sustained) the very recent proof of P != NP, my theoretical work is (while not insignificant) pretty small potatoes. Call those guys one in a million, not me.

  64. Actually I would put these others at one in several (tens?) millions, since generations of scientists, mathematicians, and scholars have worked on most of those problems at one point or another, or had the toolbox to do so, and did not solve the problem. In most cases, the toolbox necessary in those problems was smaller than this one because they are not cross-disciplinary problems (strictly speaking).

    I’ll stand by my praise, regardless of the numbers.

    Eric, I get to do what I want to do because an entire industry changed about 12 years ago and you had more to do with that than anybody else. If that hadn’t happened, I’d probably have walked away from IT entirely because I hated administering Windows boxes with a passion (and that hasn’t improved). Most of my work over that last 12 years has been with OSS, and the same goes for many thousands of others.

    I owe you a debt of gratitude that I can’t repay, so I pay it forward the best I can, but my praise is not silly fanboism. I watched what was happening, watched what you and OSS did, saw the results, listened to people on both sides attack you, read your essays and judged them on their merits.

    My own contact with the hackerdom is considerably smaller than yours, or indeed probably many others on this blog, but I’ve not run into a personality that could have pulled off what you did. If you say there are 2M hackers with the right mindset, and of those 4 with the right tool box, I’d still say half of those would not want the hassle of making themselves a target the way you did.

  65. How much of that 15 million or so lines of Linux kernel is device drivers for damned near everything that could be hooked up to a computer, and how much of that is what the average person would think of as the OS kernel?

    $ find . -name ‘*.c’ -not -wholename ‘*/drivers/*’ -exec cat ‘{}’ ‘;’ -o -name ‘*.h’ -not -wholename ‘*/drivers/*’ -exec cat ‘{}’ ‘;’ | wc -l
    3778570

    $ find . -name ‘*.c’ -exec cat ‘{}’ ‘;’ -o -name ‘*.h’ -exec cat ‘{}’ ‘;’ | wc -l
    7830209

    This only counts C code and is on the relatively old kernel (2.6.24) that I have access to at the moment, but I think it’s fair to say “about 50/50”.

  66. esr> without “ESR” we’d have had to grind for a decade, maybe fifteen years, for a fuzzier and more qualified victory

    Falsifiable?

    esr> wrong psychology – too much the classic introverted geek, not enough performer/communicator

    Eric as performer identified with by nerds only?

    esr> key insights. Somebody was going to get the rest of the way if I hadn’t; it was time

    Next major iconoclastic epiphany?

  67. Esr: ‘Some days I think something much weirder and more interesting: that there was no real luck involved, that the hacker culture created an ESR when it needed one and because it needed one. This would take us back towards inexorable tides of history and all that.’

    1. As, educated population increases we should see more and more of this.

    2. Also, an interesting note. This is what I call the opposite of a “true lie.” If everyone acts as if they believe in historical determinism then the deterministic things never come true.

    3. Since you are economically inclined, I’ll also point out that the efficient market hypothesis (weak form) is also only true if people act as if they don’t believe it is true.

  68. ESR,

    you have mentioned this Tourette-syndrome thing a couple of times, but from what I could gather about it from Wikipedia and Google (tics, coprolalia etc.) it is not at all obvious what aspects of this can be useful for public speaking or leadership skills or something of that sort. If anything, it sounds like a hindrance to overcome. So how did it help you? Or did it help you exactly because it is a hindrance and you had to develop skills to overcome it which skills proved to be useful later on? Also you have implied once that Tourette is something to be seen as some sort of a diametrical opposite of Autism/Asperger i.e. “typical geekiness”, again this is something I could not find anything about on the net.

    (BTW. Defining coprolalia as a medical condition is IMHO an excellent example of proving Thomas Szasz right that there is something deeply fucked up with the psychiatry profession. How could anything whose definition is 100% culture-dependant i.e. what is obscene in one culture or subculture is perfectly OK in another one, considered on the same level as a real biological illness and not as something one does intentionally?)

    1. >it is not at all obvious what aspects of this can be useful for public speaking or leadership skills or something of that sort.

      As I noted before, Oliver Sacks has written about people with a neurotransmitter balance shading towards Tourette’s that they show some tendency to become charismatic performer-communicator types – voluble, excitable, funny, “larger-than-life” personalities prone to jagged, creative lateral thinking and with absolutely no fear of looking foolish. Think of the comedian Robin Willams, except that he seems to have got that way by using cocaine – I remember when I first saw him doing standup thinking that it was weirdly like watching part of my own mental life become exteriorized.

      Alas, I am not as funny as Robin Williams. But the reason I’m not that funny is different from most people’s reason. Most people can’t operate in a Williams-esque mental mode at all, their brain chemistry is wrong for it. I’m like that all the time, but in normal face to face dealings with people I am a much more controlled personality than he is: the jagged lateral thinking and nonstop stream-of-consciousness jokes and big charismatic gestures stay inside. When I’m very fatigued the automatic censors relax a bit and more of that becomes exterior; I’ve never been drunk or stoned, but I suspect it would have a similar effect.

      Like Williams, I also de-inhibit a bit when I’m on stage, and this is a large part of what makes me an effective speaker. Williams has learned how to filter his Tourette’s-like mentation so the only things that come out of his mouth are funny. I have a similar trick, not fltering for funny exactly but more for what will produce intellectual and emotional arousal in the audience.

      There are other ways to be an effective speaker. If you have the right Tourette’s-like brain chemistry (elevated levels of dopamine and acetylcholine), this one is the easy way.

    2. >Also you have implied once that Tourette is something to be seen as some sort of a diametrical opposite of Autism/Asperger i.e. “typical geekiness”, again this is something I could not find anything about on the net.

      Yes. The shadow-autism/shadow-Aspergers you see in a lot of geeks produces very introverted cerebral people who don’t solve the other-minds problem well; they are poor at emotional communication and with a very few exceptions (Richard Stallman being one) have zero or negative charisma on stage. Shadow-Tourette’s cases like me are “opposite” because we’re extroverted charismatics by nature. The combination of shadow-Tourette’s with an intellectual geek forebrain is very, very rare.

      1. I wrote: “The combination of shadow-Tourette’s with an intellectual geek forebrain is very, very rare.”

        I will add that my best guess is that the most common occupation among people who fit this description is actually stand-up comic.

  69. esr: Can you recommend some reference re: HFA’s? While much of your description fits me well, the above about “other-brains”, charisma, and public speaking doesn’t (though admittedly I’ve had to exert conscious effort to overcome natural inhibitions, and I’m still personally very reserved in social settings). Perhaps I’ve just applied my geek tendencies to the problem without really realizing it. I’ve never thought of it all that much because I WANTED to overcome those enough to become a leader type as opposed to a follower or lone-wolf type (though I think my natural tendency is the latter, which I guess also points at HFA again :^).

    1. >esr: Can you recommend some reference re: HFA’s?

      I don’t think I have anything better to hand than a Google search would turn up.

  70. I’m going to indulge myself on a mini-rant that I’ve been saving for about 15 years now.

    GNU software is ‘free’ as in beer, not ‘free’ as in speech. It’s free to use the same way a ‘free’ beer is to use. It’s absolutely ‘not’ free as in a recipe to make beer that one may choose whether to make proprietary or not. Once you make the decision to incorporate GPL code into your codebase your are forever obliged to speak exactly as RMS wants you too.

  71. Well, bringing this thread back to a more personal level I can say that you, Eric, did have some impact on me personally in my early development as an IT professional.

    I had lived in Silicon Valley and was getting my feet wet in IT working for a web host as a tech sales consultant guy, helping people choose their hosting/internet services. At the time they were called Best Internet Communications and had a lot of great minds working together cultivating a UNIX FreeBSD based farm. This is where the seed of UNIX was planted in my mind so I set out to absorb as much information in book form that I could without getting bored. One day I was in a second-hand store and came across The New Hackers Dictionary which I paid less than a song for and set about reading it from cover to cover. Now I know that you didn’t write it per se, but you did a most excellent job of editing/commenting the Jargon File and putting it into an enjoyable format that I snarfed into my cranium like a hungry caterpillar preparing for the pupae stage.

    Being in Silicon Valley at the time I was exposed to the cutting edge of the WWW and programming and everything techie so it was a great experience for me and The New Hackers Dictionary helped put everything into perspective because it helped me have respect for the roots and origins of it all, it influenced me to have a sense of humor about the frustrations of computers and gave me words to put to the mind-bending implications of ones and zeros.

    I was thrilled to have been able to meet you some few months later at the very first Linux World Expo, you probably wouldn’t remember me, you were standing there in one of the hallways, I think we were all on our way to go watch Linus Torvalds speak, I came up and shook your hand and wished that I had my copy of The Yellow Book for you to sign. You were understandably distracted with all of the excitement but you were polite and cordial and then we all just went our own way to hear the great Linus speak.

    I remember many of those days in Silicon Valley with fondness and a slight bit of nostalgia when I think about the buzz around the growth of Linux and the endless possibilities we all knew would be realized with the internet and being there among people like yourself that time and working for one tech based company after another made me feel like I was part of something great.

    Today I work in the NOC for one of the widest circulated newspaper/media concerns in the world and I never forget where I came from and how people like yourself made it interesting and enjoyable for me to learn about it and helped me to get where I am today…being able to pay the bills anyway…hehe.

    I think we all need to be reminded and remind ourselves how we may have an impact on the lives of others, most specially in IT where we might help spread knowledge and truth about what it means to be a Real Hacker.

Leave a Reply to Shenpen Cancel reply

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