Bitcoin Forum

Hello, I have recently written an open-source, digital cash library with a working server and test client. I think there is much potential for integration with a project like yours.  Or at least you might see value in raiding my code for Bitcoin. 

Here are some details about my release:



WHAT IS Open Transactions ?

-- It's a solid, easy-to-use, CRYPTO and DIGITAL CASH LIBRARY.
-- Including an operational CLIENT and SERVER.
-- It's OPEN SOURCE, and encapsulates a COMPLETE PROTOCOL for transactions.
-- It's object-oriented, and written in C++ using OpenSSL.
-- With a high-level API in C++, as well as Java! (JNI interface)
-- Easy Makefiles for Mac OS X and **LINUX**
-- Featuring:
................UNTRACEABLE DIGITAL CASH (real blinded tokens.)
................SECURE NUMBERED ACCOUNTS (an account is a public key.)
................TRIPLE-SIGNED RECEIPTS
................BASKET CURRENCIES
................SIGNED XML CONTRACTS
................MANY DIGITAL INSTRUMENTS: Cheques, Cash, Vouchers, etc..

          --------------------------

ARTICLES...

Wiki:    http://github.com/FellowTraveler/Open-Transactions/wiki

Chart of instruments:
http://github.com/FellowTraveler/Open-Transactions/wiki/Instruments

Sample cheque and voucher:
http://github.com/FellowTraveler/Open-Transactions/wiki/Sample-Cheque

Untraceable digital cash (bearer certificate):
http://github.com/FellowTraveler/Open-Transactions/wiki/Sample-Cash

Sample mint:
http://github.com/FellowTraveler/Open-Transactions/wiki/Sample-Mint

Sample currency contract:
http://github.com/FellowTraveler/Open-Transactions/wiki/Sample-Currency-Contract

        --------------------------

IN OVERVIEW, THE SOFTWARE FEATURES:

-- UNTRACEABLE DIGITAL CASH: Fully implemented! Cash withdrawals of any asset
type, using Lucre. (Ben Laurie's implementation of Wagner's variant on
Chaumian blinding.) Once cash is withdrawn, the server has no way of tracking
it or linking it back to its next deposit.

REPEAT: Digital Bearer Certificates--with denominations, mints, expiring
tokens, spent token database, the works--fully-operational and ready to
integrate with your mixnets, your digital gold currency (or silver), your
anonymous network nodes, your bittorrent clients, your remailers, your secure
voip apps, your nym servers, your snazzy, new, file-sharing client, and your
censorship-resistant, distributed data store. This is what you have been
waiting for! Real, open-source, digital cash.

As well as...

-- ANONYMOUS, NUMBERED ACCOUNTS, secured by public key cryptography. Your PGP
key is your account, and the hash of it is your User ID. No other information
is stored. Each user can create an unlimited number of asset accounts, of any
type, each with its own randomly-generated ID. As long as you connect over Tor
and take other similar precautions, there's no way to connect any of those
accounts to you. (See PKTP and DMT for examples of this concept.) You can also
create as many User IDs as you wish, each with its own key, and with your
wallet software managing all your user IDs and asset accounts across multiple
transaction servers and multiple asset types.

-- It's like PGP FOR MONEY. The idea is to build this so that it supports many
cash algorithms, not just Lucre. I'd like to add Chaum's version, Brands'
version, etc. So that, just like PGP, the software should support as many of
the top algorithms as possible, and make it easy to swap them out when
necessary.

-- TRIPLE-SIGNED RECEIPTS for account-to-account transfers. This allows the
client and server to agree on balances while simultaneously not storing any
transaction history. (Client may choose to store his own transaction history.)
No money can ever be transferred or withdrawn without an authorizing signature
from the account owner. See Bill St. Clair's excellent Trubanc for an example
of this.

-- ANYONE AN ISSUER. Any user can design and issue his own currency: Simply
upload the currency contract to any server. Anyone else with a copy of that
contract can open an asset account in the new currency type. The currency
contract is just an XML file with your digital signature on it. Hashing that
file produces the currency ID, which is therefore unique and consistent across
all servers. It's impossible to change any details of the contract, including
the URL, the signature, or the public key, without entirely changing the
contract's ID. To those in the know, this means that Open Transactions
supports the 'Ricardian Contract'.

-- BASKET CURRENCIES. Open Transactions can distribute a single currency
across MULTIPLE ISSUERS. How is this possible? Users can define basket
currencies, which the server treats the same as any other currency, but which,
behind the scenes, are each simply a list of 5, 10, or 100 OTHER currency
contracts. The issuance is delegated to a basket of other currencies. Users
can easily define baskets, open accounts based on basket types, and exchange
in and out of these basket currencies using the client software. Baskets are
an important example of the distribution of risk that I believe is necessary
to make digital cash unstoppable online.

-- DISTRIBUTION OF RISK ACROSS MULTIPLE TRANSACTION SERVERS, IN MULTIPLE
JURISDICTIONS. The wallet software can store an entire list of transaction
servers. Every new server contract that you import to the wallet puts a new
server on the list. There will be many such servers, run by multiple entities
and run in multiple jurisdictions. Many will run openly and with full access
to their local court system. Others will run on anonymous networks. Users will
be able to individually choose a server for a certain transaction or even
distribute their assets transparently across a list of different servers.

-- DISTRIBUTED ACCOUNTS. Your wallet software could display a single asset
account to you, which it is actually distributing across a list of 10 or 100
servers behind the scenes, using the open transactions protocol. Of course,
you choose the servers, the number, the algorithm, etc. and it's a trust
market. In the future, this software will eventually merge into the next
generation of anonymous networks. You can be a part of it.

-- SEPARATION OF POWERS. The entities operating the servers are not actually
issuing any currencies. Meanwhile the Issuers are not operating any
transaction servers. Neither one of them is performing exchanges in or out of
the normal banking system, since those services are handled by exchange
providers in the various local jurisdictions (all separate entities.) This
provides a lot more legal legitimacy and protection to all of the entities
involved. Meanwhile all risk is distributed. The issuers distribute their risk
across multiple storage companies, and the users distribute their risk across
multiple issuers (using basket currencies) and across multiple transaction
servers (via their wallet software.)

The idea of separation of powers was first highlighted to me by Loom where
there are many different issuers on a single server. This led me to explore
further along the idea of separating and distributing risk across multiple
issuers, and multiple servers, which is what led to the development of Open
Transactions.

This product includes software developed by Ben Laurie for use in the Lucre
project.

Fascinating.  Unlike bitcoin, which is pseudonymous, OpenTransactions is claimed to be actually anonymous.  I will have to study it some more, though. 

Are you aware of any potential drawbacks or flaws or weaknesses?

The accounts are numbered (they are PGP keys) so in this respect, they are pseudonymous.

The digital cash is untraceable. Meaning, you can see that account# 38272 exchanged a token, but you don't know where it came from, and you won't know where he spends it. (Because the token was blinded when the bank first issued it.) In fact, you don't even know if he's just exchanging the same token over and over again because he's bored. To learn more about this, google about Chaumian blinding.

Then by adding an HTTPS interface, you gain full anonymity for cash token exchanges (as long as the user connects over Tor). But of course, in this case the user doesn't have access to cheques, transfers, or cashier's cheques, since he is now operating "cash-only".

Yes this is interesting. I need to do more research befor I can say anything interesting too! Kudos though!
 

The only problem with this system is that you called it digital "cash".I like everything else about it though. :)

Cash is the most accurate analogy.

When you withdraw cash from the bank, and then later someone else deposits it, the bank has a record of a withdrawal and a deposit, but they don't know that it's the SAME cash, and they don't know who had it in between, or how many hands it passed through along the way. They simply can't trace it.  These traits are also true of digital cash.

The other digital instruments are much the same as the physical equivalents:
- account transfer (a PGP-signed message, given directly to the bank, transferring funds to some other account),
- digital cheque (a signed message, given to the recipient, authorizing bank to transfer funds when presented),
- digital vouchers (like a cashier’s cheque. Funds are transferred to bank, who then issues a cheque. Bank becomes payer.),
- digital cash (Like a real cash withdrawal. Funds are transferred to bank, who then issues blinded, untraceable, bearer certificates),

Remember, physical cash is not the actual value, it is only a transfer mechanism. YOU have to trust your government that your physical cash is BACKED with real value. (Good luck.)

Digital cash is also NOT actual value, it is ONLY a transfer mechanism. YOU have to trust the issuer of that cash, and the contract he used to issue it, if it is backed with real value.  And that is why Open Transactions is designed as best possible to distribute risk across multiple issuers (through basket currencies) as well as distributing risk in other ways (across multiple transaction servers, across a diversity of jurisdictions, etc)

In Open Transactions, anyone can issue a currency type and the users can decide what they want to trade and who they want to trust.

 

I meant that governments dont like it when you mention "cash" and anonymous,untraceable pgp transactions in the same sentence :).Not that it isnt analogous to currency!

FellowTraveler:

I was hoping that more people would comment in this thread, and I'm wondering if I'm missing something.  Perhaps you could elaborate on exactly how you envision integration of your library with the Bitcoin project?  It looks to me like the two projects could be quite complementary.  I would be curious to know (since I'm not a developer for either) what features are unique to each project?  Which aspects of each project is superior?  Thanks.

I, for one, welcome our new anonymous, untraceable pgp transaction cash overlords!

I'm dubious, and here's why. You contract says:

"DIGITAL SILVER GRAMS are payable to bearer in physical silver grams, at any exchange provider supporting our contract..."

Is there one within, say, fifty miles of where I live?

"Silver Grams are audited monthly by highly trusted people."

Do I know any of them?

"Make sure this contract is enforceable in your jurisdiction."

Or it's not enforceable. Right?

"Silver and its maker, God, are solely liable to perform according to the conditions specified in this contract."

Aren't the exchange providers, or the highly trusted people liable as well?

"Any server operator using this contract will not issue any more silver digital currency than he actually has in physical silver."

But who's to stop him if he's not liable, and the contract is not enforceable in my jurisdiction?

What if you could encrypt your bitcoin transaction with pgp before you sent it out which would then be unencrypted using your key on the other end?

Yeah, I was thinking the same thing.  I think it would be pretty neat to actually send someone some bitcoins via email so that they can seed their account.  Sending coins over the social networks like Facebook would be pretty cool too.  Bottom line, I'm all for bringing Bitcoin to the masses ASAP :)

Set up a series of mybitcoin accounts and send people the login details.They dont even need a client to use it that way. :)

https://www.mybitcoin.com/open-account.php (https://www.mybitcoin.com/open-account.php)

Biggest advantage of Bitcoin, IMO:
-- Decentralized and distributed.

Unlike this, Open Transactions still has a client and a server. But I have designed it still to distribute risk as best as possible (using basket currencies to distribute risk across issuers, using wallet software to distribute accounts across multiple servers, etc.)

Ultimately, fully-decentralized and distributed is where things are going, so Open Transactions will evolve to become more and more like this. In my opinion, this will happen first on the client side. An OT client could be written that is also a Bitcoin node, and even also a Ripple node as well.  In fact, over the long time I think it will also merge with anonymous networks. I think there is a lot of potential for development on the client/node/wallet side where all these sorts of technologies will start to be tied together.

Another advantage of Bitcoin:
-- Currency backing cannot be counterfeited.

Bitcoin is curiously similar to gold -- perfectly fungible, evenly divisible, stores value in some esoteric way (so far), can't be counterfeited, can't be printed en masse by national governments...


Disadvantages of Bitcoin:
-- more traceable (then again not EVERY instrument has to be untraceable)
-- no choice of backing

With Bitcoin you can't choose to issue a currency based on gold, or based on silver, or based on a basket of dollars, bitcoins, and gold. You are restricted to one backing: computing power. This is all well and good, and in fact I think it's a really cool idea, but it also means that there will still be a demand for software that allows people to issue specific currency contracts and trade in them--especially if they have the option to do so untraceably.

As I said before, I think for my own project, the best place for Bitcoin is some sort of integration with my client side. 

What kind of value I can offer to Bitcoin? Like I said before, maybe you will find something useful in my code that you can raid for your own project. Open Transactions is an entire Crypto and Digital Cash library, so it's possible someone will have some other use for the classes in the library that I haven't even thought of. That's why I want to get it out into the hands of people like you.

Come on people. This guy is trying to introduce us to his program. Let's give him a little respect and talk about IT.

You want to get this kind of reception when we go to other projects?

Truly traveller it is veer interesting and much more up my alley than bitcoin. However there is so much more to analyze and learn than with bitcoin.

It makes one feel a little slow and behind the curve.

-----

So let's start with a noob question. Is it in use anywhere? Is there annother community forum somewhere that I can go  and lurk in?

At this point it is only a library (and server, and test client.) so it's really meant for developers only, and future prospective server operators and issuers. Looking at low-level details I can understand how complicated it might seem.

But when a nice Android client is available, or a nice iPad client, then hopefully it will not be something that requires much analysis at all. It should be something that Grandma can use.  She can add new currencies to her wallet, and new ID cards (pgp keys), and she can open new accounts in any currency type, and she can write cheques on those accounts, and she can transfer to other accounts directly, and she can even withdraw in digital cash form, or get a cashier's cheque.  She'll have some Mac app for doing it with drag and drop.

Blinded cash issuers (using for example this system) and the bitcoin system could make a pretty good combination.  It could be similar to how private banks kept gold reserves and issued bank notes:

* Use bitcoin or a modified bitcoin/RPOW/bitgold like system to create a securely auditable, unforgeable, public, non-anonymous "high power money" analogous to gold.

* Issuers (or "banks") of securely anonymous (blinded) bank notes keep bitcoins as reserves.  By analyzing their public bitcoin chains, any customer can audit the reserves of any of these issuers.

Those who don't care about their privacy or who want to be issuers with trustworthy reserves can use bitcoins directly, while cash customers who like their privacy can use the securely anonymous blinded cash.   That forms a two-tier system, analogous to the old privately issued money system that involved mining and storing gold (bitcoins) and issuing bank notes backed by gold (anonymous cash backed by bitcoin reserves).

The whole system could be far more secure than the old gold-reserve + banknote system, because the "gold" is more difficult to steal but far easier to securely audit, and the "bank notes" far harder to counterfeit and (especially against modern investigation techniques) more difficult to trace/identify users.   The "gold window" for these banks could be far more secure from various threats than in the old private bank note issuers.  And it's all conveniently online.

So my take is to take advantage of the public and basically non-anonymous nature of bitcoin/RPOW type systems to securely audit people who claim to own X of them (such as currency issuers), while using blinded digital cash for transactions where privacy is more important than public audit.

Thanks for an excellent post.You should post a bitcoin address to receive hash cookies for it  :)One thing is for sure - banks today dont let you go into their vault and see their reserves easily.I wonder how much gold is left in fort knox ? The answer is either zero or it is fool's gold - gold plated tungsten ?

This is quite insightful and obvious in retrospect. Bravo.

This is all predicated on the actual "computing power backing" being able to retain "intrinsic value" as a drop-in replacement for gold. But if that turns out to be the case (e.g. if the BitCoin continues to rise against the dollar, or at least hold steady) then this would work. I'll be curious to see how it all turns out.

Hi FellowTraveler,

I love your solutions. I have been archeticing much of the same for a while and just not got around to coding just yet.  You may have saved us all bunch of time and given us a head start  :)

I have sent you a private message you should still check.

Meanwhile,  I understand your ideas for multiple server nodes where people can chose and manage where their value is held at from their wallet.  But what is your solutions for redundant data storage in the case where a  server/banking/contract node fails or worse, confiscated unjustly.   Do your server nodes support the redundant data exchange between them?  Ideally the server nodes should communicate through a decentralized P2P backend like bitcoin as that is one of the advnatanges and attractions of it.  Such a redundant data backend would account for worse case senarios or to work around the lose of any node. 

In normal day to day operations maybe they can exist in some for of hierarchy to provide faster clearing of transactions which a P2P network could take time to clear under certain conditions/senarios.  With designated 1st, 2nd and 3rd tier server nodes in each jurisdiction or that are first tier for a particular basket/currency/contract issue.  Not too much unlike the way root DNS servers work (but this is one way downstream) and High Availablity Systems work.   If one server node does go missing or down,  the server/P2P nodes can reorganize and designate a new one.    Possibly a hybrid of an automated P2P network with independent trusted human node operators.

Thoughts?

On the clienet side...
I would be interested in creating a web client coded in python client.... if you can provide information on how I can tie python into any of your java or C  libraries.

Advantages of Open Transactions that bitcoin maybe can not provide...
OT can reach other markets faster leading to faster adoption.
1) Allowing entities to issue currencies can make it easier for local communities to use adopt; issuing their own locally controlled and backed and local agreed upon currency while at the same time tieing their system into what could be a more global network of server nodes.
2) Allowing faster adoption by people that do not understand bitcoins.  You do not have to EXPLAIN a gold or silver coin to anyone living on earth.  Thus a metal/commodity backed and issued currency by a recognized and reputable issuer can gain in populartiy faster as well.  LR and Pencuix are great since you can own gold/silver through them but I do not trust their centralized operations.  But if a popular silver mining company issued and backed silver grams into a decentralized  OT system, you could count me in.  Local merchants would be more willing to accept such digital silver more easily as compared to convincing them to accept bitcoins.

Anyone with bitcoins could also issue bit coins into the OT network which indeed has a value in its rarity alone.

As others touched on in this thread.   The crediblity of the issuer ideally needs to be proven satisfactorily to the end holder/user of the currency.  There are a number of ways to do this... of which I can share my ideas later.

I would also be interested in making public statistics available on each currency which the server nodes can accumulate.
Volume of transactions, amount in circulation, # users holding a particular currency, etc... This could ultimately lead to a rating system as to the most widely used, liquid and trustable currency issues.

Ultimately, the entire system would function much like the stock and future markets allowing the trading and transfer of assets between individuals/entities (alternative real open market) ultimately putting an end to all the behind the scenes coruption in todays  FX, futures and equity markets.    Putting control back into the PEOPLEs hands and away from the few elite controllers that have a monopoly on money and exchange today.... forcing us to play by their rules where as people now days have the code and power to take back such control in the form of bitcoin and OT like systems.

Regards,
Xen

Are there any precompiled windows builds ?

Did anybody ping satoshi on this thread?

Open Transactions must be centralized, so it doesn't fulfill the goal of Bitcoin.

Two very important paragraphs, thanks for this great idea.  DigitalCash + Bitcoin = PGMS = Pretty Good Monetary System

I like the idea of OT cash backed by bitcoins.

I suppose you can run an OT server on a TOR hidden server? Throwing TOR into the equation gives us a highly resilient system which that government can do little to shut down. Awesome!

Philosophical Tangent...
It seems that the growth of the internet and social networks is quickly out-pacing the ability of governments for control them. I think we are going to see some revolutionary shit in the next 10 years. Bitcoin is one example.

I think that neural networks are going to be a huge thing. The internet will soon become I giant AI hive mind of neural networks and human integration with this system will become more sophisticated with direct brain-to-computer interfaces. We are becoming the internet; the next phase of evolution. cool shit! :-)

I have this idea called "protocol AI" in my novel. The idea is that business interaction are mediated by a network of AI whom jobs is to learn reputation, make judgement calls, and so on. They are manned by game theory experts, who are called judges, but their real job is to make their protocol AI the best mediator on the internet.

It would be trivial for government(s) to shut down Tor's directory servers. There are only a handful of them. I think that I2P has a far better design in this regard. It has no central directory servers, the use of UDP (as well as TCP), streaming support, and the ability to TX/RX over different paths. Not to mention that I2P's hidden services (called eepsites) are WAY faster than Tor's hidden services. :D :D

The auto update feature is the only scary part of I2P. If the authors were forced to trojan, or otherwise tamper with the I2P update, the network would be screwed. This can be remedied by switching the update feature off and manually auditing each update of course. :)

Go one step further: artificial neural networks are probably going to surpass the capacity of their biological counter parts (humans). Eventually we'll have ANNs designing ANNs! Your Judges could themselves be computers.

Sounds like a cool book. Have you read Necromancer? It explores the idea of purely digital individuals.

This idea is really cool, I just repeat it below to expand it a little bit.

1) An OT issuer backs up his "OTBitcoins" with bitcoins. These "OTBitcoins" provide instant payment and untraceable digital cash.
2) Everybody can verify that he does not issue more "OTBitcoins" than he really holds, simply by inspecting the public bitcoin blockchain. This is something that could be done automatically in a combined bitcoin OT client whenever a user wants to purchase "OTBitcoins"

Now here comes a little expansion, I would be interested what you think of it:

There is one little problem in the above example:
The OT issuer can still be unable to deliver bitcoins if he lost the private key to the bitcoins he used to backing up the "OTBitcoins".
So just inspecting the public blockchain will not help.
The OT issuer of "OTBitcoins" also has to proove that he has access to the bitcoins which form the backup.

Solution:
The OT issuer backs up his "OTBitcoins" with two bitcoin addresses.
He then sends *continuously* back and forth bitcoins between these two addresses, so that everybody can see that he really has access to the private key and that he can make good his "OTBitcoins" with real bitcoins.
Performing this check could also be done automatically in a combined bitcoin OT client whenever a user wants to purchase "OTBitcoins".

So at least at the date of purchase of "OTBitcoins" the user knows that the issuer can make up good the claim of delivering real bitcoins.

Of course if the issuer looses the private key in the future or runs away, the "OTBitcoins" from this issuer will become worthless.
So the idea of multiple issuers forming a "OTBitcoin" basket currency will help to reduce this risk.

Since I very much like the idea of crowd-based services, the amount of "OTBitcoin" issuers could be equal to the number of bitcoin users.
The only thing needed for it would be that a combined bitcoin OT client allows the user to become a "OTBitcoin" issuer with a part or all of his bitcoins in his wallet. The parts of the bitcoins used for issuing "OTBitcoins" would then start the above ping-pong sending back and forth of bitcoins between addresses and would be locked from normal bitcoin spending.
As a reward the user receives a share of the "OTBitcoins" equal to the bitcoins he dedicated to this from his wallet.
He can then use these "OTBitcoins" for instantanteous payments and untraceable payments.

In a nutshell, every owner of bitcoins can sign up to become part of a crowd-based "OTBitcoin" bank which provides untraceable digital cash.
I would like to see how governments try to shut down or control something like this!

Maybe there are better or easier solutions.
Curious what you think of it.