The point I was trying to make on the thread I started about a global
distributed datastore is that I thought modeling the system in terms of
probabilities is a better approach than trying to enforce a systemwide
currency that gives exact tit-for-tat measurements.
When you don't have any guarantees of SLA when you "purchase" things with
some virtual currency, what is it actually worth? I think a better approach
is to cryptographically log the behavior of various participants in the
system in order to automatically make judgments about them and how reliable
you expect them to be.
On Mon, Aug 22, 2011 at 3:15 PM, Jon Cox <j...@experiments.com> wrote:
>
> Dear p2p-hackers,
>
> After thinking about the similarities between the commons issues
> faced by the PaulGardner-Stephen's Serval project & Zooko's Tahoe
> LAFS, I was motivated try (yet again!) to refine my understanding
> of currency systems, barter and money.
>
> I'd like some help!
>
> These concepts seem to have direct bearing on problems that arise
> when bootstrapping new users into a p2p/mesh system, trust, credit,
> fairness protocols & so on, yet I lack a standard terminology for
> talking about them in a precise way.
>
> First, I'd like to share a speculation:
>
> My guess is that it would be best if Tahoe LAFS computed all
> credits and debits in terms of its native "commodity currencies"
> (this term is defined below) like storage, availability, bandwidth,
> latency, and priority, and then possibly mapped these things
> to other currencies via forward contracts in a fairly pluggable
> manner.
>
> While Bitcoins might be interesting as one of several possible
> side-channels to establish "credit" for the system's own native
> currencies when dealing with non-boostrapped strangers, or for
> those who would otherwise hit their "debt ceiling", I'm hoping
> that friends could still offer credits for "native" commodities
> denominated as such *directly*, without needing to think about
> a fluctuating 3rd-party / non-native currency abstraction.
>
>
> That said, here's my first shot at defining a few terms, along with
> a note to the Serval project that may be of more general interest:
>
>
> o Benefit
> That which produces a net increase
> in some desired state of being.
>
> o Intrinsic value
> Non-bartered perceived benefit.
>
> o Value
> Optimally bartered perceived benefit.
>
> o Money
> The mathematical abstraction of value
> Note: this can be a positive or negative quantity.
>
> o Currency
> The concrete manifestation of money.
> Note: currency always has a non-negative value.
>
> Example: physical dollars & coins.
>
> o Pure liability
> Something of negative value that isn't the result
> of owing anybody something.
>
> Example: an inadequate reputation
>
> o Debt
> A liability resulting from owing something of
> positive value to another party.
>
> o Pure asset
> Positive value that isn't the result of someone owing you.
>
> Example: good health.
>
> o Credit
> A liability that another party owes to you and recognizes.
>
> o Commodity currency
> Some fungible good or service that is easy measurable
> and comparable, transferable and transportable,
> sufficiently divisible and durable, widely bartered
> with known rates of exchange, and derives its value
> from its intrinsic usefulness rather than its role
> as a currency or as an item of speculation predicated
> on the existence of a greater fool.
>
> Example: cigarettes in a POW camp.
>
>
> o Native commodity currency
> A commodity currency whose production and consumption
> are intrinsic to the economy itself.
>
> Example: Carpool rides in a ride-sharing network.
>
>
> o Collectible currency
> Like a commodity currency, except that its value comes
> from the mutual speculation of those who create demand
> for it, rather than its intrinsic worth to any end user.
>
> Example 1:
> Bitcoin is a perfect example of a collectible currency
> because its intrinsic worth is exactly zero (you can't
> even use them to line the bottom of a bird cage).
>
>
> Example 2:
> Gold and silver are best thought of as being somewhere
> between commodity and collectible currencies; however,
> the volatility of silver prices relative to its supply
> and industrial demand shows it's more on the speculation-
> driven "collectible" side if things.
>
>
> o Fiat obligation currency
> Like a collectible currency, except that every unit created
> represents the transfer of value from someone who has actually
> produced a good or service of non-zero intrinsic worth to a
> person or institution that has offered nothing in exchange
> for it but the currency itself. Crucially, the production
> of fiat obligation currency is restricted by law, and demand
> for it is created by requiring its use. Hence, every unit
> represents a claim by its producer to simply extract things
> of actual value from those who cannot create the currency
> themselves, and yet are bound to use it by law or necessity.
>
> Example:
> The deceptively named "Fed", a group of privately owned
> for-profit banks, create legal tender (US dollars) out
> of thin air, then "exchange" them for US Treasury for
> T-bills. The T-bills have real value in that they represent
> fractional ownership of the US government's ability to
> extract goods and services from its own citizens, and from
> the citizens of foreign countries. To do this at home, it
> puses the IRS, federal marshals, and the legal system.
> Abroad, it uses military and/or political power to enforce
> its will directly, or it can use intermediaries in its thrall,
> such as the IMF, The World Bank, various resource-rich or
> strategic client governments, and so-called "Coalitions of
> the Willing". Hence, every dollar represents the assertion
> of a one-sided obligation; they aren't coupons for anything
> of intrinsic value that the Fed used to barter with the
> US Treasury in exchange for a fraction of the tribute our
> government is able to demand. Instead, they are more like
> souvenirs commemorating an outright confiscation of it that
> has already taken place. The obscene material wealth and
> power held by those happy few who are on the receiving end
> of this arrangement is the direct result of the compulsory
> exchange of intrinsically valuable goods and services for
> inherently valueless slips of paper decorated with stars,
> eagles, and the faces of dead presidents.
>
>
> o Fiat commonwealth currency
> Like a fiat obligation currency, except that the value
> extracted in the process of its creation goes to the community's
> own public fund, rather than being siphoned off by a private
> aristocracy.
>
> Example (I think):
> Treasury-issued "United States Notes".
>
>
> -----------------------------------------------------------------
> Here are some comments I made to the Serval project recently.
> I'd love to get some feedback on them.
>
> Hopefully, the ideas I'm tossing around are of wide enough
> applicability to merit general interest and/or or an
> informed & corrective critique
> -----------------------------------------------------------------
>
>
> Serval already has at least three commodities with universal value,
> and native non-speculative demand: bandwidth, latency, and priority.
> Using these in combination with a system of reputation and credit as
> a "commodity currency" makes a lot more sense than dragging in a
> collectible currency.
>
> Assuming nodes in a Serval mesh are free to associate in different
> virtual communities of reputation (as humans do in real life),
> all a community needs to do is restrict mesh access ("the commons")
> in ways it sees fit to discourage behavior it dislikes, or grant
> special credit for positive actions (eg: donating bandwidth,
> particularly in times of shortage). It seems as though you could
> even model the policies of Serval "community" as an autonomous
> System (AS) on the Internet, and then go on to think about
> communities honoring each other's credits with bandwidth, latency,
> and priority as parameters like a AS-AS forex. This would keep every
> community free from central domination by other groups, yet open to
> engage in mutually beneficial data transfers. It seems likely that
> a few huge communities would naturally arise, along with lots of
> little ones. Users might like to have different policies for each
> of them, as different levels of generosity (or blind trust) might be
> appropriate; one might have reputations in personal groups, affinity,
> regional/global, etc. It seems like the core requirement is the
> ability to assign a policy profile to a group ID, and the ability
> to have more than one.
>
> If someone with a sufficiently large balance and community reputation
> never gets bumped by somebody who might be free-riding or has no
> reputation, it makes sense to build a positive and credible
> "accounts receivable" bandwidth balance. Someone could always cheat
> you with a series of bogus IDs, if doing so entails them being
> second-class citizens the entire time, there isn't much incentive.
>
> I think it's good to place the main focus on first degree contacts,
> as has been done with great success in hawala networks. The hawala
> system is well known for its efficiency and real-world security, even
> over wide geographical areas lacking any common authority sanctioned
> to use powers of state (such as garnishment) in order to enforce
> contracts. Reputation-based access to the commons can go a long way
> to regulate behavior. Diamond traders that operate via handshake
> agreements are another nice example of regulation by community
> restriction. More generally, the principles and practices of
> Islamic banking look like they're worthy of serious study.
> I've only begun to explore this topic, but the prohibition against
> outright interest seems to have generated a fascinating array of
> partnership-oriented contracts and non-leveraged financial
> relationships -- all with a bare minimum of bookkeeping.
>
> If an adversary's goal is to hunt down all copies of a message and
> destroy it, another consideration emerges: perservering ignorance.
>
> Example: Julian Assange has no idea how to reclaim all
> copies of insurance.aes256; therefore, he cannot
> be forced to help anybody else do so either, even
> under the threat of torture.
>
> Had a recording been made of every person who
> downloaded the file, then thugs could go after
> these people too, even if they number in the
> tens of thousands. However, what if these
> folks placed the same data on the net in a
> way that couldn't be detected, and was then
> downloaded by yet another set of people
> that's unknown to the first group.
>
> The same logic applies to the release of the
> key to unlock the encrypted data. You probably
> want some sort of anonymous quorum where the
> various members don't actually know all the
> other members, but share an ability to look
> at a common "outer envelope" wrapping a sig
> and an inner blob. The disconnected cliques
> have these blob-bearing coconuts floating
> between them, yet the inner blob part within
> the coconut cannot be cracked solo by any
> one clique; however portions of the key
> to crack it can be posted publicly by them
> (in any form: wrapped by the group envelope
> or not, steganographically or not, signed
> or not, and so on).
>
>
>
>
> Cheers,
> -Jon
> _______________________________________________
> p2p-hackers mailing list
> p2p-hackers@lists.zooko.com
> http://lists.zooko.com/mailman/listinfo/p2p-hackers
>
--
Tony Arcieri
_______________________________________________
p2p-hackers mailing list
p2p-hackers@lists.zooko.com
http://lists.zooko.com/mailman/listinfo/p2p-hackers
