Geodesic Markets, CyberDog, and Internet Finance

Robert Hettinga
Shipwright Development Corporation
44 Farquhar Street, Boston, MA 02131



I'd like to talk about why we need to teach a new dog, Apple's OpenDoc internet superset project called CyberDog, a few old tricks. Reminds me of the "Wallace and Grommit" episode where Grommit is reading "Electronics for Dogs", so he can help Wallace build a moon rocket in the basement. I believe they were after cheese. Anyway, I'm interested in teaching CyberDog how to spend and collect money, digital cash in particular.

A little while ago, an "essay" of mine blasted out of my fingers one Sunday morning. It was about what I called geodesic software, an idea lifted heavily from Peter Huber's geodesic network idea of 10 years ago. Huber said in a report to the Feds that exponentially collapsing microprocessor prices had dissolved the old hierarchical phone network into a geodesic one, like the geodesic structure of Bucky Fuller's geodesic domes.

In a geodesic network, there is no top or bottom, the path from any node to another is random, and any given path through the network depends on the network's load. We'll be bumping into Peter Huber again a little later on.

It's easy to us to see now that whatever order exists in a geodesic network like the internet is emergent, much like crystals emerge from molecular structure, or speciation emerges from random mutation, or conciousness emerges from random brain cells firing.

The "essay" I wrote, a posting to the apple-internet-users e-mail list called "The Geodesic Network, OpenDoc, and CyberDog", eventually will make it in a necessarily truncated form to InfoWorld. The editor told me when he read it, it reminded him of a scene from Tom Wolfe's _The_Electric_Kool-Aide_Acid_Test. Something about Neal Cassidy turning around from the wheel to face the passengers, in the middle of a double-clutch on a steep mountain road, to talk about Zen while driving Ken Kesey's famous day-glo school bus named "Furthur".

I think I must have made him nervous. Thinking that I could calm him down a bit, I told him that while I had read all of Mr. Wolfe's work, I really liked Hunter Thompson a little better, that I didn't own a car, that if I did, it would be a Hummer with a gun rack or something, and that it wouldn't have "Furthur" painted across the top, but it might have a "Windows 95 = Macintosh 89" bumpersticker on it. I don't think he calmed down, but he immediately changed the subject...

The point of my little rant was that it had dawned on me, between bites of a complementary ham and cheese sandwich at an Apple product demo at this summer's MacWorld Boston, that compound document architectures like OpenDoc's enabled geodesic software to exist, much in the same way the collapse of microprocessor prices enabled the geodesic network to exist. OpenDoc breaks up all the code necessary to get a certain job done into smaller and smaller randomly associated bits, like a drop of dish soap does to globs of grease in dishwater. These bits of code can be anywhere from another part of my hard disk to somewhere on the other side of the world.

The beauty of OpenDoc is that it is software out of control. Whatever structure you get from a collection of OpenDoc parts is emergent, it is not planned. My page full of OpenDoc parts doesn't have to look, or work, like yours. This flexibility is what attracted me to the Mac to begin with, and it's great to see the same thing happening to what has become a hyperfeatured application wasteland.

By the way, of the best books to look at on emergent systems like the internet and other biological systems ;-), is Kevin Kelly's aptly titled "Out of Control". Kevin Kelly is one of Stewart Brand's people, who in turn was one of Ken Kesey's people, those Merry Pranksters of the Neal Cassidy, bus-named-Furthur, fame. While Kevin's now the chief spinner of mind-candy at Wired magazine, don't hold it against him too much. We'll bump into Mr. Kelley again a little later here, too.

It seems to me that the new, improved, 90's version of Minot's "vast wasteland", the microcomputer desktop, is dominated by Microsoft, who is now attempting to grow a kudzu-like monolithic application superset, using, paradoxically, their compound application architecture, OLE. In OLE, you can't call a small feature out of an application without loading the whole damn app, much less calling a small, autonomous "part", which does only one thing very well. If it's not loaded already, you would have to call all of, say, Microsoft Word 6.0 to edit a little text frame, and of course, you could die of old age waiting for that to come up. When I'm feeling particularly malicious, I think OLE is the biggest boon for memory manufacturers since the invention of Windows.

Finally, I said in my rant that Microsoft, playing dog in the manger like this, could end up choking on all that code. Remember, in a geodesic network, one node cannot possibly process the traffic of the entire network. I think in an environment of compound documents, the same can be said for piece of code in a software geodesic, in this case an application like Word 6. A geodesic network routes around all obstructions.

After the "essay" went out, I got a call from an evangelist at Apple named Ford Johnson who had read it and said that the same kinds of things had occurred to him. He saw a hardware implication in this as well. Netscape, SGI, Oracle and others believe client-server architecture to be the end result of all of this. Servers are the new mainframes. Everyone will have dumber and dumber "client" machines on their desktops. Of course, it's is more likely that at the outermost edges of the internet "clients" will get smarter and smarter, and eventually the distinction between clients and servers will disappear. Most machines will be their own POP, News, and WWW servers. This is a good thing for Apple, which has always preferred a peer-to-peer network architecture anyway. With something like OpenDoc's CyberDog surfacting away all that concentrated information, the process would accelerate. Client-server is just a first-order effect of a fractal process, the creation of geodesic software.

Which brings me to how I got to thinking about all this.

When I first moved to Boston in late 1986, I had a "nameless" phone number in my pocket. Where I had moved here from, Chicago, that wasn't a problem. Since the 1930's Illinois Bell had a service called CNA, for Customer Name and Address. You call the CNA number, gave them your nameless phone number, and if it was listed, they gave you the name in the phone book associated with the number. This isn't free, they charge you 35 cents extra for the call. So, I try to do the same in Boston, and I call New England Telephone information. The NET operator thinks I'm from Mars. If what I'm asking for isn't illegal, it should be. So, chalking it up to more regulatory ignorance in hidebound New England, I hang up. A little later that day I get yet another brainstorm, though. I jump back on the phone and I call Illinois Bell, and I talk to the supervisor of the CNA operators, who kicks me upstairs, ad infinitum, until I'm talking to a staff analyst to a vice president who's responsible for the marketing of the phone number database at Ameritech, Illinois Bell's parent.

I suggest to her that I'd like to clone CNA in Boston on the end of a third-party 976- prefix number, and she laughs. 976 numbers had just been tarriffed in Boston within the past year, and they were the local equivalent of 900 numbers today, and in those days, 900 numbers were used only for polling, and not for selling stuff like they are today. People were just starting to agitate for commercial, interactive, 900 numbers, and we all know what happened next, but I'm just trying to see if I can set up a cross-reference directory in Boston like they have in Chicago, because obviously NET doesn't have one.

So, when the analyst stops laughing at me, she says that since Judge Green has broken up the Bell System, CNA has been declared what's called an "enhanced service" according to the decree breaking up AT&T. The Regional Bell Operating Companies (RBOCs) are prohibited from offering "enhanced" services like cross-reference directories, but that Illinois Bell's CNA is grandfathered. No other Bell company can do it, and NYNEX, New England Telephone's parent, had already been slapped down once for offering a similar service in New York called "Hello Yellow", a cross-referenced yellow pages directory.

"So, what's the problem?" I said. She says that while I can offer the service, and I could get nearly identical lists of independently compiled Boston phone numbers, and that Ameritech would be more than happy to help me technologically, I should expect massive resistance from NYNEX, whose network I was going to put this service on, and who was going to be collecting the money for me from my customers. That was because this guy Peter Huber had just come out with this report to Judge Green called "The Geodesic Network", saying, effectively, "Deregulate them all, and let God sort them out." I talked about this before, so I won't waste more of your time on that, except to say that I went down to the U.S. Government Printing Office and I got Huber's report.

Reading this report got me to thinking some more. I decided I would write this cross-directory idea up, because it was interesting, and, by writing it up, I might be able to find someone to weasel some investment money out of. This was right before the market crashed, and there were lots of those folks around in those days... But in writing this up, I Saw The Future -- for the first time, anyway.

My "future" wasn't quite on the John Campbell continuum, but it was astounding nonetheless. The thing that got me going was the idea of pricing. If you could "toggle" the price of a phone call at will, then you could sell anything you could say or do on a phone line, for any negotiated price whatever, and the phone company could settle your trade.

I knew all about settling trades, because the only real job I could get after not graduating with an undergraduate philosophy degree from Missouri was as a clerk in Morgan Stanley's Chicago operations department. I wasn't a very good clerk, and I now know why they call it "the cage", but I did learn all kinds of things about how the capital markets work from the underside of the rock where all the crawly stuff lives. Things like what clearing houses do, for instance, which is what I saw the phone company doing, in this model of information "trading" with variable-priced 900-style phone lines.

Like I said before, we all know what happened next. 900 numbers took off, mostly for phone sex, and the stock market crashed. While the erst-info-economy I saw coming was cool to think about, I still had to pay the rent. So, I went back to building static business models for people with dynamic business processes, automating management reports, and other fun stuff people thought they needed a squad of COBOL guys for, all with a little Mac SE.

For a while, I worked on contract with Fidelity, and then I got a real job with Citicorp, and when Citicorp pulled the plug on the operation and handed me a nice severence check, I went back to contracting, knowing now that I made much more money not knowing where my next dollar came from.

I kept doing little jobs for big corporations until a few years ago, when I invented something I thought was pretty neat. It was a rather large server-based Frontier script we called ProjectServer. Scott Lawton of PreFab Software built the script, and Joel Bowers of JMBA built the FileMaker front-end and server. ProjectServer linked MacProject and FileMaker Pro, using AppleEvents. People could see a multi-user FileMaker list of the stuff they had to do, across all the projects they were in, and they could report their status on those tasks by, say, checking off a check box when they were done with it. Whenever a task changed in FileMaker, ProjectServer would come along at night and update the task in a project manager's MacProject file. To start a project, all a project manager had to do was to drag his new MacProject file into a "blessed" folder on ProjectServer's host machine. With folder aliases and File Sharing, this was just a matter of dropping the file into a folder. ProjectServer would update the FileMaker database with any new MacProject files or changes in old files.

Polaroid, who we built this for, let us have the code. On a trip to California, I demoed ProjectServer to Dennis Marshall and John Perr, the Claris product managers for FileMaker and MacProject, and they offered to let us stuff their software boxes with a blurb for ProjectServer. I was psyched, and left with visions of software moguldom dancing in my head.

Needless to say that didn't happen. I'm living proof that you can't start a software company on a consultant's income -- much less in a garage -- anymore. It seems like everytime I turned around, it cost money. Money incorporating Shipwright to own the code. Money to pay for the 4-color blurbs for the Claris software boxes. Money in phone charges for a fax blitz to all the Claris Solutions Alliance members. Money to show up at trade shows. Money to pay people who knew little about what we were doing to write press releases to people who knew even less, so the customers, who actually knew something, could read about it in the trade press. Money, Money, Money. We sold exactly one copy, and gave a bunch more away. In the end, another company with $10 million in revenue, who was spending 10 times as much as ProjectServer cost to develop on their advertising budget alone, built an identical product for the Windows market (Microsoft Project to Access). The embarassing thing was, they used the name "ProjectServer", and didn't know we were even out there.

After all that, I still believe the major reason we didn't make it with ProjectServer was mostly my inexperience. I wasn't really cut out to be a Software Mogul.

I was, as the art director for my ads said, an "Idea Hamster". Which is why I now write stuff like this, and try to earn my keep on the schmooze circuit. If and when what I'm doing gets big enough, I'll just hire a business manager so I can just keep that hamster-wheel spinning.

Meanwhile, there wasn't much money left to mess around with ProjectServer anymore, and, one night, sitting in Boston's South Station waiting for a train home, I read one of the first issues of Kevin Kelly's Wired magazine. There's an article about how an MTV VJ had taken his Mac IIci and put it on the internet with the domain name of He said, "As soon as you can upload cash, the record companies are out of business." Bingo. This is how to sell ProjectServer, I thought. On the internet.

The reason I thought this was even possible was that I had seen this before, with Huber's Geodesic Network idea, but also with the financial industry as a whole. In banking and in the capital markets, it's called disintermediation. Mutual fund companies are first-order creatures of this process. In short, technology, the very phone switches which were making the telephone networks more geodesic, were removing the middlemen, the network of distribution for capital and the orders trading that capital. That's because the capital markets are an information hierarchy, like most of industrial society.

I have written in detail about this elsewhere, but it's easy to map the growth of large distribution and manufacturing hierarchies to the falling cost of communications. Sears, Roebuck was built on Rural Free Delivery, you could mail a catalog with the offer, customers could mail you orders bidding for products. The growth of centralized stock and bond markets exploded because of the telegraph, and later the telephone. Economies of scale are a function of the ability to gather and distribute information and resources from a central place, usually stacked on a hierarchical network of wholesalers and distributors.

Huber pointed out that the structure of the phone system was hierarchical because lines were cheaper than nodes. Lines are still mostly wires, but nodes were originally people: the telephone operators. In the 1920's the phone company had to start automating switching, because at the rate of growth of the telephone system, they would have eventually had to hire everybody in the country to switch calls. Mechanical switches begat electromechanical, electromechanical begat electronic, and electronic begat microprocessor switches. Remember, Shockley, the inventor of the transistor, invented it to make faster phone switches.

At each point the cost of switches, that is, the nodes in the network, fell some very large amount, and, with the advent of microprocessors, this cost implosion has become the central economic fact of the modern era. It even has a "law", an observation, really, that we all like to talk about. It's called Moore's law, which says that the price of a given microprocessor falls by half every year. When Gordon Moore of Intel saw it first, in the 1960's, it was "only" half every 18 months, so if anything, the trend is accelerating, for the moment. Nodes are now cheaper than lines, and they're going to get way cheaper. The folks at Xerox PARC have been watching this for years and have been playing with the technological consequences: they call it ubiquitous computing. They talk about how your alarm clock will say "Coffee?", and if you say "Yes", it will tell the espresso to fire itself up.

It is this phenomenon that has collapsed the phone network, and it is collapsing every information hierarchy we know, which means most of our industrial social structures.

Much has been written lately about the implosion of hierarchy all around us. People swarm around in amorphous ad-hoc work teams. Middle management is becoming scarce. More and more production is drop- shipped from the factory in response to mail orders, to phone orders, and now to web-orders. At some point, factories will probably get closer to their customers as they get smarter, faster, and smaller. On the other hand, this proliferation of microprocessors has increased tremendously our ability to "micro-ship" quantity-one orders using overnight freight, making warehouses and other fore ward staging areas smaller and more scarce. It's hard to figure out exactly which way things are going, except that monolithic production is a thing of a past, along with monolithic anything else.

This is not to say that the concept of hierarchy will disappear. After all, social mammals live in hierarchies, or at least a pecking order of some kind. Dolphins "banish" other dolphins from their company. Status is very important in baboons, chimps, and in what we could find of "anarchic" hunter-gatherer societies at the end of the industrial revolution. However, it seems to me that the gigantic control hierarchies we have built in everything from business, government, religion, and even in recreation in the past 150 years or so will start to "surfact" into smaller and smaller entities.

So now that we're getting what looks like a geodesic economy in the physical universe, how do we get one in the mother of all geodesic networks, the internet? That's where we pick up my story again.

When I got home, after reading that Wired article about the MTV VJ, I logged on to NEXIS, -- which is a great boon when I can afford it -- and I did a search for articles containing "internet" and "bank" and "cash". One of the first "hits" I got was an article for the Whole Earth Review by, guess who, Kevin Kelly, about money, and cryptography, on the internet. It's an article about David Chaum, who has invented a protocol for digital cash, about Whitfield Diffie and public key cryptography, and about Tim May and Eric Hughes, and the group they started called the Cypherpunks, an anarchic band of people who meet once a month in the Bay Area to talk about cryptography and who have an internet mail group to do the same. It seemed that Eric, at least, wanted to start a bank on the internet which underwrote digital cash.

I hadn't been on the net since I left the University of Chicago, where I had gone to finish school by transferring credit to Missouri, and to learn things like math, stats, and computers. The last thing I did there was work from midnight to eight in the morning, running big production jobs until they finished or broke, and fixing them or waking up the programmer if they did. Nasty boss, great job. I learned to pump all my online sessions and Usenet news and mail into EMACS windows, with alarms on all of them for when things broke, to smoke cigars the size of baseball bats, and to play blues harp, all while eating large pizzas with everything on them. I even got pretty large myself as a result.

Anyway, this digital money stuff got me back on the internet. Fortunately, it was then possible to get a commercial internet feed, so when a friend at Apple turned me onto an Adam Engst book called "The Internet Starter Kit", I called The Internet Access Company in Boston, and they set up the domain for me.

The first thing I did, once I figured out how to play with all the new toys, was to find, and sign myself up for, the cypherpunks e-mail group. Needless to say, the world's a little different now, after only a year and a half.

Just using software I've loaded off of the net, I know now how the world is going to work with an economy overlaid on a geodesic public network using strong cryptography. I know how to send messages which even the NSA won't be able to read for decades, Moore's law and all. I know how to sign e-mail messages so that people know it's really me. I know how to send perfectly anonymous messages. I can buy things and only the seller knows what I've purchased or for how much.

I know that in the future, it won't be your reputation, it will be the reputation of your public key, or even several of your public keys, if you want. With just a clerk's knowledge of the capital markets, I can see how it will be possible for people to trade digital bearer certificates in complete personal anonymity, certificates representing any known financial instrument, without the need for a central, hierarchical securities exchange, or even a private, closed, and proprietary network like the NASDAQ system. I can even see how giant mutual fund companies -- the most famous result of financial disintermediation, the effect of Moore's Law on the financial markets - -- will be surfacted into smaller bits just like their predecessors were, those financial giants like the Morgan companies.

I know how to use totally anonymous digital cash, yet cash so secure that if you double-spend it, you are identified at the bank for doing it. Digital cash in denominations so small that it's cost effective to do transactions in hundredths of a cent.

There are a lot of problems with all this. I like to say that things are fraught with opportunity right now. People talk about how anonymity isn't all that it's cracked up to be, with talk about perfect crimes, and, of course, hysteria about the "Four Horsemen of the Infoclypse", as Tim May calls them: Terrorists, Drug Dealers, Pedophiles, and Tax Evaders. All of whom are going to use strong cryptography, digital cash and perfect anonymity to (with apologies to Mr. Carlin) Warp Your Mind, Curve Your Spine, and Loose the War for the Allies. Of course, cypherpunks and electronic privacy advocates know that this is just FUD (Fear, Uncertainty, and Doubt), and as Mac users, you know all about FUD in the hands of your opponents.

Meanwhile, there are people on the other side of the argument, myself included for the most part, who think that unprotected computer networks have an enormous capacity for information abuse by all sorts of people, including corporate database marketers and other hackers, not to mention just plain criminals. That information abuse could extend to destroying personal freedom as the surveillance capabilities of nation-states increase.

Of course, there's the counter argument that in a geodesic network, no single node can monitor everything. That's called "security through obscurity", and while it's not quite hiding your head in the sand, it's asymptotically close. Fortunately, just hanging out with the schooling fish to prevent a bad case of shark bite is not necessary. Strong cryptography is very strong. Modern encryption algorithms depend upon the mathematical inability of computers to be fast enough to factor very large prime numbers and thus break a given cryptographic key. There's a wealth of literature about this already, in particular Bruce Schneier's "Applied Cryptography", and the discussions of this stuff are all over the net, and I point you to those sources if you want to learn more or argue about the particulars. Because, of course, I have bigger fish to fry here.

That's because I liken strong cryptography to aviation. Some people just had to fly. They literally killed themselves figuring out how to do it. To them, it was an inherent good. They had ideas about what they might do when they could fly, but mostly they just had to do it. However, the reason aviation is commercially successful now is because flying is the fastest way to get anywhere. People aren't buying the majesty of flight, they're buying coach fare to Cleveland.

The same thing goes with strong cryptography, and, even more so with anonymous digital certificates, especially digital cash. Cryptography and anonymity is security from prying malefactors, which is an inherently useful. There are sharks out there, and some of them even wear suits and claim what they're doing to you is for your own good. But cryptography is also the key to commerce on the net, not just to keep people from sniffing your credit card off a router somewhere, but to allow the creation of peer-to-peer digital certificate-based transactions, in particular, cash transactions.

Digital certificates and cash are the ultimate economic surfactants. They allow you to build a unique digital bearer certificate pointing to anything of value, including corporate debt and equity, derivatives as complex as the market will bear, and finally, cash in a bank somewhere, and to trade, to clear and to settle that trade for next to nothing in transaction costs.

I have symbol that I use to describe a US dollar on the net, "e$". I say e-money, or e-dollars, depending on how I'm using it, for instance e$20.00 is "twenty e-dollars", but e$ by itself is "e-money", with apologies to those mark, pound, and yen users out there. I truly understand that there are other economic units out there besides dollars, but at the moment, the US dollar one of maybe 4 currencies of record, and it makes a good proxy for all money, which, of course is all fungible by definition anyway.

On cypherpunks and elsewhere, I usually flag my posts about e$ like I've flagged this one, with an "e$:". That makes it good filter- fodder, and I'm also sure there more than a few kill-files out there with that flag in them by now ;-).

We're in the home stretch, now. The most thing to take home from all of this is the following: Internet commerce is financial cryptography.

Without a cryptographic financial protocol, you can make a trade, but you can't clear or settle it. For instance, if you're feeling reckless, you could just punch your credit card information into an unsecure Netscape web-form. However, the trade still has to clear, that means that the book entries are swapped on proprietary and secure financial networks linking the vendor, the credit card company, and yours and the vendor's banks. Once the trades are acknowledged, the trade is settled by a bank wire, usually through a clearing house, usually the federal reserve system. There's a secondary transaction that occurs when you pay it all of in your monthly VISA bills, and that's when the whole trade finally settles, but we'll ignore that for the time being.

Now, let's take the other extreme, buying some Ben & Jerry's Cherry Garcia from the local convenience store. You walk in, you get your Cherry Garcia, you pay for it with cash, you walk out. You have just executed, cleared, and settled a trade all at once. That's because both you and the store owner trust the currency you handed her. It stands on its own as a store [ahem...] of value. The owner can turn around put that cash in the bank and immediately (or at least as fast as today's bank books can be closed) cut a check to Ben & Jerry's local wholesaler for more ice cream. The grocer doesn't need to know who you are, or where you got the cash from. It costs her much less in that transaction's overhead to deal in cash. They don't say "Cash is King" for nothing. The smaller the transaction, the more it needs to be in cash. The smaller the seller is economically, the more his business needs to be in cash. Otherwise, we'd be using our credit cards in vending machines, or to kick in on the office football pool.

Interpersonal transactions are almost all in cash, or cash equivalents like traveler's or cashier's checks or money orders. The last time you sold an old car, did you take MasterCard?

The good news is that the technology exists to do cash transactions, down to the hundredth of a penny. You can buy software directly from the guy in Singapore who wrote it. You can download music directly from a garage-band's web server. You can get legal or medical advice. You may be able to buy teleoperated labor. All of this, and pay for it in cash.

Also, as software becomes more granular, more geodesic, and as it operates on insecure geodesic public networks, it's going to have to pay it's own way, and yours. That's not just because copyright and patents are a joke on a geodesic network, where it's impossible to enforce claims against anonymous customers and where a watermark only tells you where the first pirated copy was stolen from, it's because that's the most efficient way for the originators of new software and new information to get paid.

Now, lots of people fantasize about how someday software will be metered, and most secure schemes are hardware-based, with a glorified dongle counting the use of an app and then narking to the developer at periodic intervals or just croaking the code until the magic incantations made after crossing the developer's palm with silver, but imagine a world where the code just sends you money, no matter where it is.

That's right. Your code sends you money. If everyone has digital cash on their machine, and the code is geodesic, then an OpenDoc part, say, would just request ask the digital cash wallet to send some smidgeon of that cash somewhere, and upon getting a digitally signed receipt, for that payment, an operating certificate from the developer, if you will, the code would then get to work. You could put all kinds of parameters on the payment stream, like a certain number of free executions, or payment for a certain number of executions, like rent, or whatever. The limiting factor, of course, is the market and your competitors. Your code had better have friendly payment terms, or it had better be damn good, and totally new, because your competition simply has to reverse engineer your code (no way to enforce copyright or patents, remember?), and woo your customers away from you. You code could even find out if it is the most recent version and upgrade itself.

You've seen all this before when people talked about software agents, and all the wonderful stuff they could do. Now, with OpenDoc and the idea of geodesic software, it's possible to see how all software, not just a special class of interpreted code, could operate in this chaotic, emergent fashion.

More to the point, having a geodesic software microeconomy is probably the best way to provoke software evolution. If you think of digital cash as software "energy", then you have a perfect, certificate-based accounting system for all things digitable. No need for external audit trails because the certificates (operating certificates, digital cash) stand alone. It's easy to see how with these measurement tools, it could then be possible to actually see the benefits of software, services, and the use of information in quantifiable terms, instead of just guessing, which is what current industrial accounting forces you to do.

So, we've finished. What we call digital (I hate "electronic"; why neglect the poor neglected photon?) commerce is really financial crytography. Financial Cryptography is an economic surfactant, just like Moore's law on the internet is an information surfactant, and compound document architectures such as OpenDoc is are software surfactants. This is why we need to give CyberDog, OpenDoc on the internet, one of Grommit's favorite books: "Financial Cryptography for Dogs". All we need to do then stand back and watch the fur fly.



Bob Hettinga


[ Back to the previous Rant ] | [  home page ] | [ On to the next Rant ]