Announced by Evgeni Pandurski:
"I am currently working on a P2P system for extended bartering. I almost have finished the design of the system and have done a toy implementation of one of the subsystems. I plan not to release any code in public until I have most of the design frozen. I believe this is going to prevent the project to be spoiled down by too much design and philosophical discussions. Anyway I am open to discuss technical details with anyone who is interested in the concrete design and implementation.
Let me try to briefly explain the general idea:
The biggest problem with money is that it basically introduces artificial scarcity of one particular good - the money itself. All the other problems are a logical consequence from the mentioned problem - interest, fiat money, credit. Barter exchange does not have this problem because it does not introduce any scarcity that have not existed before. The only problem with barter is that it does not works when a wide variety of goods are traded.
This is called "double coincidence of wants" which spoils things down. So basically barter would be nice if it worked at all (but it does not).
I propose a kind of generalization of barter which I call "Circular Exchange Trading System" or CETS. The ordinary barter is a special case of circular exchange of goods where there are only 2 participant in the exchange. General idea is that lots of people declare what they have for sale and what they seek to buy and the computerized system offers them appropriate circular barter transaction. Details are not important at this level of abstraction.
What is important here is that "double coincidence of wants" problem disappears because the guy you deliver what you have and the guy you obtain what you need form are different.
If the thoughts above are true. Is seems that such a system once implemented and widely accepted would free exchange from any dependency of centralized institutions (or at least lessen this dependency greatly). The previous statement is obviously too enthusiastic. It is very hard for me to deduce all the drawbacks this approach to trading (CETS) certainly has. I would be glad to hear from you all possible critiques of the idea I presented to you."
You have an important point here - once implemented and widely accepted - and I believe that is the point the various LETS schemes fall down on. They can't seem to get the critical mass needed to make the exchange work as a generalized system.
Money works because it is anonymous. It does not care what has been exchanged for it before - it is always ready (interest and investors allowing) to finance something new, and on a personal level it is always ready to buy whatever you want to buy.
The drawback with LETS (and you bring that right into your system) is that finding something to exchange is kind of cumbersome. Both ends of the exchange have to actually do something extra before the exchange becomes possible. One has to go to the database and offer his/her wares. The other has to go to the database and choose what he or she wants to buy. More often than not, the buyer can't find all he/she wants among the offers that are being made. This is the real limitation - the bottleneck.
So if the system is going to be widely accepted, that bottleneck has to be taken care of. The marketplace has to be somehow expanded. And the best way I see to expand the marketplace is to make the currency (or system) independent of any central database that collects offers and allows deals to be made by matching the participants. This function has to be pushed out to the periphery in some way, for a system to become viral and spread enough to become a real player in economic exchange.