Control Key Scroll Wheel
If the text on this page is too small or too large, hold down the CTRL key on your keyboard and move the scroll wheel of your mouse back and forth. The text should grow or shrink. If you don't have a mouse with a scroll wheel, hold down the CTRL key and press the pl­us or minus keys. This works with most browsers and many Windows based programs including Microsoft Word.

Building an Operating System for the Year 2025

This idea folds into a theory of mine concerning “monopoly behaviors”. These behaviors exist throughout human interactions. They are merely attempts at gaining advantages, and knowingly or unknowingly almost everyone indulges in them. One would be a fool not to. The heavyweight champion of boxing leverages any advantage possible in order to remain champion. You or I would do the same to keep our jobs.

To me a monopoly behavior is any attempt to hinder a legitimate experiment from taking place because a positive outcome would not be to ones advantage - or to inject extraneous elements into such experiments. It's as simple as throwing a non sequitur into the mix to confuse a discussion taking place at work. These behaviors stop innovation in its tracks. Although such behaviors help individuals in the short run, and for individual lives there really isn't much else than the short run, they create vast inefficiencies in systems, perhaps dooming them in the long run. What follows is how the creative process could work, stripped of such interference.

Building an operating system for the year 2025

First fact: When I worked at IBM on operating system source code, to the very right of each line of code were the initials of the person who wrote that line. Sometimes the original code would be patched because there was a problem with it, a bug, and even the fixing code could have fixing code to it, if there were something wrong with that.

The code might be improved and new function added to it. Replacement code, and a notation of the line number where that code applies, was kept in a separate file from the base code. These fix, or patch files were dated, so the compiler knew which patch had to applied and in what order. In this way one could reconstruct the entire history of a piece of code or module, and in theory figure out who wrote every line of every version of an operating system.

Second fact: There are things called digital certificates, also called digital passports or public-key certificates. These allow encrypted messages to be sent via email and establish the exact time when an email is sent and received, and by whom. Everyone should be using this method to send email at all times - but in fact I can't even talk software designers into taking a few minutes to get their free digital certificates, so that we can encrypt emails when we exchange ideas and insure they won't be stolen.

Third fact: Cheap disk space is potentially the memory for us all. By all, I mean all of humanity. Storage media has become so cheap, and the processors so cheap and fast to search that data, enough disk space exists to remember everything anyone has ever thought or dreamt, with version numbers and dates for each idea and dream. (Hyperbole alert: Don't take this last statement literally.)

Fourth fact: Any complex task can be broken up into a succinct list of subtasks. The most efficient way to turn out cheapest best quality products is to define those tasks and let out each on an open bid basis. Anyone can enter, the winner judged by an independent as-unbiased-as-possible panel of ones peers.

This leads me to believe we could build a virtual corporation from the ground up, whose purpose is to design, implement, and market a single piece of software. This is the quickest, cheapest, and most efficient manner to create the PC operating system that finally sends Microsoft Windows down to the foul dark realms from whence it sprang.

The software will be designed, coded, debugged, and sold, with profits distributed based on the value of an individual's actual contribution, estimated only after the product is fundamentally complete, in the marketplace, producing revenue. This is as opposed to the convoluted process of negotiation to make sure that one gets everything one can possibly get, right up front, before anything is actually produced - an enormously expensive ritual in terms of overhead costs and disastrous to human creativity.

Participants email in their ideas, submitting their solutions for accomplishing a specific task. This input goes into a moderated forum. Digital certificates place the exact time emails come in, resolving disputes as to who proposed an idea first. (These certificates are already legally binding.)

Moderators aren't allowed to submit solutions to their own forums but receive a share of profits, based on how good a job they did finding best solutions. But moderators should not know the identities of participants nor participants of each other. So that if one is building a macro for the I/O subsystem, the code must stand alone. Nothing about who one is or who one knows should be allowed to influence the choice of best way to do the job. Past successes do not count, only current output. This is the ideal, but I always define things 'fuzzily', by which I mean that within the set of real solutions for any task or problem, one can only pick the best one that works.

The overall scheme works because software is the pure case. You needn't be any specific place to write or submit it. So with competition open to anyone with an Internet connection, the most talented people in a given technical area could be working together.

Everything is open. Suggestions can be viewed immediately (after being logged). One person can suggest changes to another person's solution. These permutations of an original solution become solutions in their own right. If you think someone else has a better idea, abandon your own and grab a piece of that one.

If I have a good solution there's no reason my competition must compete by only using entirely new solutions. Everyone piggybacks on what comes before. Immediately when an idea comes into the public realm (after being logged) it becomes fuel for further ideas. When one person suggests changes to another person's idea, these permutations become solutions in their own right. If you think someone else has a better idea, abandon your own and grab a piece of that one. Seeing things so, one begins to understand why anything that takes a whole loaf approach, all or nothing, stinks of monopoly.

Many different individuals can receive payment for devising the winning solution to a given task. Just putting forth conclusive reasons for choosing the best solution merits a taste of the honey. We would have to come up with an algorithm weighing relative values for all contributors, plus some way to fend off the hoard of lawyers representing people who claim they got a raw deal. I leave those tasks to anyone who wants a shot at it.

Any idea can be shot down. Aggressive individuals often reach the top not by providing the best possible, but rather by providing the winning solutions to problems. It is in their interest to attack other solutions.

Because of this IBM nearly went bankrupt in the early 1990's. Management stuck with mainframe hardware revenues as a primary source of income, while for over a decade people in the industry knew that software would be preeminent.

IBM was controlled by people who knew how to make and sell hardware. This was their solution. Focusing on selling software would mean giving up control, so rather than step down and make way for change these captains let the ship sink with them.

I am not going to be self-righteous about this. It is senseless to counsel individuals to go against their own best interests. Only saints or idiots will do so. This is a standard human behavior and standard human behaviors exist for a reason. We must create and publicly publish an agreed upon rule set, then find a way to enforce that set of rules, or as close as humanely possible. If this is possible it is only now possible, using computers and the Internet.

Few ideas emerge fully formed. You can't find the value of an idea until you look down the various paths it could take, its permutations. Nor can one fairly reject an idea unless you follow down these paths. In my scheme one has to let a solution not work. But if you show why it can work, or can alter it in such a way that it works, you get a piece of the pie too.

I believe that it's better to have many people receive a fair share for creating a superlative product, rather than a few people receiving huge compensation by controlling the process. Value-add is a critical concept. If someone gives advice that helps a project along, they taste the honey too. Thus everyone is invested in making a solution work, or honestly finding out as quickly as possible why it won't.

We can map everything virtually, all inputs and outputs. Everything is seen immediately, from design to coding phase to debugging. When a piece of code goes on the web, programmers will test and debug it immediately, knowing they will receive fair compensation for their contribution. No one need take on a whole task, just what they have time, energy, desire and ability for. This is applying excess capacity most efficiently, as judged by the actual people who possess the talent, not by a management team standing above, telling them what to do.

The process is open, much like the Linux operating system. But they give Linux away for free. I don't believe in that. Why should the best people dedicate their time and energy to something which has no chance for material rewards? Here we take the very best people and turn them into entrepreneurs in the areas they know best.

What marketing scheme do we follow? However the moderated forum decides. The same goes for legal work. No one gets paid up front, nothing by the hour. You receive a share if and only if your solution wins, or you have made a contribution to the winner. Let the best idea fly. Let each person do what they do best and be judged by their peers. I'm not sure how this would work but I'm sure someone can come up with an idea.

Once the code goes out to customers it stays under the same scrutiny and competition. So long as my solution is still employed every month I get my check. If someone else comes along with a better solution, theirs supplants mine in the next version. They get a check, but so do I, if for no other reason then for having provided a target to shoot at. So a value is placed on solutions that are most easily replaceable. Under our current system I would never advise someone to follow this strategy. It is to ones benefit to make oneself as irreplaceable as possible and clever people find all sorts of interesting ways to do so. But that is not our system. Using our system, how long will it take to create software that sends Microsoft Windows back to the dark realms where it belongs? Give us six months. That should be enough.


© 2001 by Alan Silverman. All Rights Reserved.
May be distributed for non-commercial purposes with attribution to author.

Home Page Writing Trustimonials Right Way

Email me

If the above link doesn't bring up your email program, copy and paste this address into an email: