March 21, 2004
The Digital Silk Road
DSR is a historical pre-commercial (circa 1994) shared accounting architecture that was proposed to compensate router owners for passing the packets of other entities.
Cooperating router nodes would count packets passed between them, and occasionally, they would send "number" money packets back and forth to reset the counters. These paid-for resets would cause charges to trickle across to big users, and money towards working routers. Defences against cheating/fraud were limited to signed notifications of balances and a simple payment system.
DSR is like LETS for routers. As a thought experiment in multi-agent accounting, it is interesting for its influence on later micropayment systems (Mojo Nation?), but it assumes pre-commercial net-style honest behaviour and the absence of competition. It also suffers somewhat from the cool engineering approach ("the silk road was so cool, let's rebuild it") that always gets steamrollered by markets and marketing.
E.g., FedEx beats the original silk road, as does a host of other transport innovations such as trains, bulk container ships and blind men with canes. In today's Internet world, large corporations achieve internal Coasian efficiencies by owning thousands of routers and not doing internal charging, but collecting flat fees from customers.
Posted by iang at March 21, 2004 07:35 AM
Tamerlan and other Mongol rulers before him made the Silk Road safe from bandits. The first thing the great Khans did was implement a system of paper money as the only form of commercial payment. So the Silk Road was one that was created over a series of wars. The reason the Mongols marched and invaded was because of a great famine in their own land. Mongolia is experiencing the same type of famine today.
The Mongols, in need of fresh land due to famine, stumbled upon the more civilized nations of Central Asia and conquered them. The interesting thing is the Mongols kept the civilizations whole and adopted what they found worthy of keeping. So the earlier system of payments between routers was probably worthwhile saving après the invasion of the Mongols.
The Internet was developed in the environment of affluence and civilization, and we are now being invaded by Mongols who may or may not displace the bandits. No matter how it winds up the period of affluence and civilization is over, and payments systems must be worthy of retaining, or suffer the new ruler (whoever that may be) finding it unacceptable. The Internet is very similar to the Central Asian nations prior to the Golden Hordes' invasion of small kingdoms of civilization, sitting there waiting to be overrun.
The Silk Road was much more than a simple transportation system. I have worked and lived in Central Asia and had an office on part of the Silk Road in Almaty. What you find is that culture blends between east and west, between nomad and city dweller. These places only blossom with peace, affluence, and civilization; the Internet is a place like this. Since this last round of civilized effort is over, i.e., the era of Pax Americani, the world, like the Internet, waits to see what will happen next.
The Internet is of a civilized nature and tends to be weak when it comes to proactive measures needed to stop invading hordes. Here comes a dark age one where no trade happens and cities build walls rather than roads to other cities.
Ian you're loathesome! <grin>
DSR might have been nonviable 10 years ago. Today, computing power and bandwidth are thousands of times cheaper. The market price of 802.11b wireless routers is around $20 now. Do you all understand what's inside those? 11 megabit WiFi wireless access point with a range of hundreds of meters, or over a mile with external antenna. Bridge from the wireless universe to the completely different flow control of ethernet. Plus, a 4-port ethernet switch, and unix host computer with filtering rules, NAT host, DHCP server, webserver for configuration, etc. That's a lot of computing for $20.
However, capacity is not infinite and the critical element missing from a large-scale urban mesh is only the lack of social agreement how to measure or charge each other for transit across our hub. SPAM is a great example of when happens when no effort is made to charge for use of a resource. SPAMMERS send 1 million ads for viagra, or 1 billion, what's the difference?
PS you say competition won this battle? I am now paying over $60/month for DSL and the speed is 64 up 384 down. I am also paying over $60/year for 3rd party hosting, domain names, etc since the proliferation of hackers makes it impossible at home. Ten years ago I was paying $20/month for 28K. Monopolists won this battle, not competition.
Doh! Todd, it's *nothing* to do with the power of the node. We were running full scale payment systems that could do DSR back in 1996 on "ancient" Sun IPCs at 12 MHz! Even then we went away from "fast crypto" to "real slow Java" in preference.... Give me more power and more bandwidth and my response is <YAWN....>
On your comment that we lack a social agreement, that's correct. This however, does not mean that such a social agreement when found is cost effective, or even exists! Sure, SPAM exists and makes us gnash our teeth, but the alternate must defeat all comers.
DSR does not. The monopolists, or the oligopolists won. Sorry about that, get used to it, its economic reality.
PS: can't you go back to dialup?
My understanding of DSR is that it was *not* "number money" to reset, but "here we use some as-yet-undeployed secure money" to reset.
The interesting parts of DSR are (a) the unauthenticated bilateral accounting, i.e. before the reset, that we adopted for Mojo Nation and that was used in the PPay paper. (b) the whole notion of long haul transport via a chain of local deals -- the "Silk Road".