A history lesson from Bill Joy

A history lesson from Bill Joy

Summary: Personally I expect Linux on Cell to blow Wintel into history -but history is behavioral,and if we don't change, it'll just repeat on us, right?

SHARE:
IBM doesn't offer its own Linux distribution. Nevertheless they've already shown the first Cell based blade server running Linux, and I expect the full line up, from desktop to data center, will too.

With that context in mind, I thought something Bill Joy said in a 1999 interview with Eugene Kim of Linux Magazine. might be thought provoking.

Please bear in mind, however, that he's not generally known as someone who makes negative comments -but while he would probably like to forget the frustrations leading to this one, that's part of what's thought provoking about what he said.

Here's how the exchange went:

LM: Why do you think that Unix was never successful on the desktop?

BJ: Because Microsoft had a person who was very greedy and who was very brutal in his business dealings and was handed a monopoly by IBM due to ineptness. They had several opportunities to rein this guy in and the management blew it. So the IBM monopoly got transferred basically due to blunders. Microsoft is a direct successor to the IBM mainframe monopoly. The corporate guys coalesced around the PC standard because it came from IBM. Not because it was any good.

Personally I expect Linux on Cell to blow Wintel into history -but history is behavioral, and if we don't change, it'll just repeat on us, right?

 

Topics: IBM, Hardware, Linux, Microsoft, Open Source, Operating Systems, Servers, Software

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.

Talkback

15 comments
Log in or register to join the discussion
  • Bill Joy = the BETTER Bill

    Bill Joy is totally correct in that quote. But transferring the monopoly probably SAVED IBM - since their own anti-trust case fizzled out soon after. IBM did the RIGHT thing and learned from its mistakes. Duh!

    "Personally I expect Linux on Cell to blow Wintel into history -but history is behavioral, and if we don?t change, it?ll just repeat on us, right?"

    I see Windoze in a long downward slide into obscurity (emphasis on LONG). *NIX will prevail not just over Windoze, but all other OSs (Right now that role is Linux - which is taking away market share from proprietary UNIX solutions. The reason is that UNIX SHOULD have been on top, but its ambition was thwarted by forking, and Linux represents the "good" things about *NIX - without the fork).

    InHell will be MUCH more difficult to defeat than M$. They have superior technology (65nm process up and running today), and superior designs (the Itanium IS superior design - the P4 is NOT EVEN CLOSE). Re-hashing a RISC chip with multiple cores (Cell), will NOT match up with EPIC - as long as they share the same nm process i.e. RISC requires MORE chip real estate than EPIC. Couple this with IBMs ability to deliver (ask Apple about that), and you have a LONG climb to topple InHell.

    Bill Joy should be talking with Apple to sell some SPARC chips to.
    Roger Ramjet
    • You're not my evil twin, right?

      Maybe a long lost cousin? Anyway, you're right -except that I see Intel as the loser in the CPU wars (more on this next week) and their announced intent to produce multi-core Itaniums as oxymoronic given the nature of long instruction set computing.
      murph_z
      • Long instructions

        "their announced intent to produce multi-core Itaniums as oxymoronic given the nature of long instruction set computing."

        EPIC is evolutionary technology. Just as RISC made the instruction set less complex, EPIC makes the pipelining/branch-prediction stuff less complex - by shifting it to software. YES I know that software development lags hardware. YES I know that the first generation EPIC compilers sucked. YES I know that HP is not the greatest development platform. But EPIC HAD to happen - make the hardware LESS complex, and add USEFUL things like more registers and ALUs.

        BTW. All Itaniums are multi-core, if you count the x86 compatability "core". I've heard it doesn't work so well, but its there.
        Roger Ramjet
    • What might have been.

      Unix had a long lead over Microsoft, but lost it. Forking was part of the problem, the smaller part. By far the bigger issue was ignoring user friendliness and making life especially easy for developers.

      Now Microsoft is defeating Unix where they compete, and Linux and other open source projects are eating into Unix from the inside.

      We're seeing the long slow slide of Unix into obscurity, not Microsoft.

      And Microsoft will never have to worry about Linux, because it doesn't exist as a commercial force. Yes, companies like IBM exploit it to sell other products, and that's why it's helping beat Unix into the ground.

      To beat Microsoft, a competitor has to be like Microsoft because Microsoft has shown what causes people to buy software. Such a competitor, well-funded, patient, and knowledgeable, might have arisen. But thanks to the way open source has eaten the potential demand, that competitor will never arise.

      Linux and the rest of open source have assured Microsoft's increasing dominance far into the future. I'm not happy about that, but it's a truth to be recognized.

      Open source is the Unix community's final blunder.
      Anton Philidor
      • Careful there Anton, you are forgetting a few things

        Linux is a better deal for a hardware company than Unix as there is an ever expanding community to provide free product support. Whereas companies such as IBM used to have to use R&D money on both the hardware side and the OS side, now the balance can go towards their 'sweet spot', the hardware.

        And the amount of value returned to the open source community by any of these companies is a pittance as compared to the value received by them - but that is the devils bargain that has to be made by the OS community.
        quietLee
        • Devil's bargain

          IBM does have a Linux install and service company. It's called RedHat. And, to hedge their bets, they put substantial money into Novell's SuSE. Control without legal responsibility.
          So IBM is not especially concerned about "free product support" from a community.

          Also, corporations are not going to wait for a post on a community message board to provide service. They want professionals, and are willing to pay for them.
          They, too, are not interested in "free product support."

          And IBM contributes paid programmers to the Linux (for example) effort. Makes certain that the work they want is done and done well. Other companies collaborate.
          Even free labor is not the most essential component any more.

          The cheap price is key. Gets IBM through the door. When customers want more capabilities, IBM has the expensive stuff to sell them.

          IBM will also support the software they provide, even though they're not legally responsible for it. That's why the references to "IBM Linux" in the EC official's speech awhile ago.


          The operant word about Linux for IBM is: cheap.
          But not free.


          The best part of your post is the Devil's bargain.

          The hobbyists and philosophers who started Linux et al were not getting wide distribution.

          Then IBM promised to vouch for and contribute to the efforts that interested them.

          And in return those who had been in charge gave up the significant portion of control. They sold their souls.

          At least Faust received knowledge and a beautiful woman. I think open sourcers sold cheap.
          Anton Philidor
          • Altruism

            The open source community is altruistic - they give WITHOUT expecting anything in return. When IBM comes along and uses OSS, you say that OSS programmers "sold their soul", since IBM profits from their altruism.

            I am truely sorry (not really) that commercial programmers have lost jobs because of OSS. If their talents were SO important, then why would some of them give it away for free? Its hard to understand altruism and charity, but to many people its IMPORTANT.
            Roger Ramjet
          • But that is perfectly fine - no argument there

            I frankly don't care what anyone else does with his/her life, talents or lack thereof. If it doesn't put a dime in my pocket, it's irrelevant.

            Your right to starve or make riches ends at my well being - and I extend to you and all the same right and consideration.

            Altruism is nonsense, and is not evident in anything I've seen or read about the OSS movement.

            Getting something for nothing is an economic benefit.

            Selling a service that supports or creates an open source service or that pukey over used phrase 'solution' is an economic benefit.

            Working on an open source project, and then using that project in a resume to 'get a paying gig' is an economic benefit.
            quietLee
          • When all is said and done tho...

            I will stipulate to the facts as you have laid out, but I don't think I outlined the context fully as to my comments.

            To whit - MS wholly owns and provides internal and external support to the NT kernal. MCSE's provide support of MS products - and the substantial majority of them are not compensated by MS for their work.

            In contrast, the Linux kernal is a work in progress, not owned by anyone as a financial construct, and is freely available to use, change, alter etc. as long as the agreed upon rules are observed. While there are Red Hat/Linux engineers - they are comparable to MCSE's in that they are independent agents.

            The difference is - MS owns their product, and is able to exercise rights over it (such as mandating use, and receiving compensation). IBM also does this in terms of their proprietary mainframe OS's - as does Sun with Solaris and HP with HP-Unix.

            What I so quickly stated was that IBM et. al get he ability of OS support (not integration of product or setup, they do that themselves) of Linux without having to pay for it - an all gravy situation for any hardware company. I am not talking about logging into a newsgroup and asking a question - those issues they are able to handle themselves. But that staff would exist irregardless as IBM etc. would need that staff to support their own clients. That cost is a wash.

            What I was trying to say, in fewer words, was the cost of the business in having to create, maintain, and better their own proprietary OS. That is a cost they do not have to bear with Linux - therefore it's a far better deal for them. And that while they contribute to the community, the value they get for using the product far exceeds the value they return to the community. Again, a great deal for them.

            None of this is either good or bad, it's simple business.
            quietLee
      • Antonability

        I don't see M$ "increasing dominance" in any way, shape or form. They are ALREADY dominant in many areas, but >>I<< see that dominance DECREASING i.e. The IE share of the market slipping. When you're on top, there's only ONE way to go!
        Roger Ramjet
        • Different markets

          You're right that eventually Microsoft reaches a share over 90% of a market and stalls. At that point, the competition can't be reduced further.

          Then Microsoft switches to a different market.

          Look at servers.
          Microsoft hasn't been competing for long, and has had some quality problems. But in a significant milestone ZDNet didn't include, I think, Microsoft recently outsold Unix for the first time.
          Yes, Linux is growing also. But mostly at the expense of Unix. I think ZDNet was wrong to say in an article the entire Unix loss came from Sun, but they are recognizing the movement.
          When Microsoft exceeds the Unix/Linux combination for the first time, that will also be a significant result.

          Then think of media, XBox, small computers of various types, and a number of other areas. People have enjoyed Microsoft's slow starts in some of these areas and gloated when they decided to pull out of some. But they are well on their way to success in a number of new areas.

          Microsoft thinks of their potential market as everywhere software is used. That was part of their defense in the anti-trust case. Tried to persuade the Judge the desktop was a minor part of the market.
          In that context, they think that they do in fact have only one direction to go: Up. Barring effective competition, I wouldn't bet against them.
          Anton Philidor
      • There's UNIX, Unix and *nix

        UNIX, Unix and *nix

        "UNIX" referring to the proprietary UNIX systems (SCO UNIX, AIX,
        Solaris, etc), "*nix" referring to the unix-like or unix-based
        systems (Mac OS X, open sourced BSD derivatives, Linux, etc),
        and "Unix" commonly being used for both.

        "Unix had a long lead over Microsoft, but lost it. Forking was
        part of the problem, the smaller part. By far the bigger issue was
        ignoring user friendliness and making life especially easy for
        developers."

        You ignored the number issue with UNIX, the AT&T antitrust
        settlement the excluded it selling the UNIX OS. Without a single
        entity marketing the system, forking was a natural progression.

        "Now Microsoft is defeating Unix where they compete, and Linux
        and other open source projects are eating into Unix from the
        inside."

        Now I assume you are referring to UNIX systems. The UNIX
        market is shifting thanks to continued processor improvements
        and cheaper open source alternatives. I not sure which market
        you're referring to where MS is defeating *nix though. Yes the
        market for windows server is growing, but so is *nix (and at a
        greater rate than windows).

        "We're seeing the long slow slide of Unix into obscurity, not
        Microsoft."

        Maybe UNIX (except in the dwindling share for big-iron servers),
        but not *nix.

        "And Microsoft will never have to worry about Linux, because it
        doesn't exist as a commercial force."

        No? Guess they'll have to revise their SEC filings;-)

        "To beat Microsoft, a competitor has to be like Microsoft
        because Microsoft has shown what causes people to buy
        software. Such a competitor, well-funded, patient, and
        knowledgeable, might have arisen. But thanks to the way open
        source has eaten the potential demand, that competitor will
        never arise."

        You obviously only see a single company being capable of
        competing with MS, I strongly disagree. If the EC continues
        through with its antitrust ruling, MS will lose its only weapon
        against competition: restricted access to its desktop monopoly.

        Applications is all that holds back Linux on the desktop. We live
        in interesting times.
        Richard Flude
    • Yes & intels ability to deliver performance w/out huge backside cache

      which is why this chip is expensive & will take more than
      another year before it gets both cheaper & more efficient.

      IBM will prevail.
      johnpall@...
      • You didn't read

        my response to closely. IBM has had trouble keeping up with demand for chips. That's why Apple is shopping around - IBM doesn't deliver.

        Bulding fabs is an incredibly expensive undertaking! With InHell ALREADY making 65nm chips (I don't think IBM has this yet), it bodes well for them to maintain a technological lead. InHell could ALWAYS build and market chips well - it was their technical designs that sucked.
        Roger Ramjet
  • Rule bending

    I am glad in some way that Bill G bent rules and started the whole PC market. Marketing is the key word today. M$ has marketing power and that is what keeps it alive. X86 has been a great proving ground for *nix as a PC product. Linux is much bigger than the little engine that could. More like a friendly giant using it's power for good and not evil.
    xstep