On a decentralized bitcoin-based stock market...

[ripper234]

I wonder if there's a way to propagate this bit of knowledge so misinformation about this stops spreading.

Does it matter? For most intents and purposes this is just a technical implementation issue. For normal use it doesn't really have anonymity implications either, it's known that all funds in a given address belong to the same entity, so it doesn't matter which of the outputs they choose to use to send. Only when you want to do fancy stuff like we've discussed here it matters.

I agree it's not relevant for 99.9% of Bitcoin users. Oh well.

[Red Emerald]

There isn't any reason why this project can't work along side GLBSE. In fact, I think that would help make it much more successful.

The genesis block for our chain could contain everything in GLBSE at whatever time we decide to launch.  When the chain gets launched, a new GLBSE back-end would also be launched that would use our chain instead of w/e database it is using right now.

[jtimon]

There isn't any reason why this project can't work along side GLBSE. In fact, I think that would help make it much more successful.

The genesis block for our chain could contain everything in GLBSE at whatever time we decide to launch.  When the chain gets launched, a new GLBSE back-end would also be launched that would use our chain instead of w/e database it is using right now.

Maybe they prefer to stay the way they are. Remember that the trades will be slower if they're executed within a chain.
But they can always do both.

[btc_artist]

There's no reason there can't be more than one stock exchange.

[DeathAndTaxes]

There's no reason there can't be more than one stock exchange.

Which is why I think a hybrid solution may be the best bet.

Have the stock ownership and transfers handled by a decentralized blockchain and then allow multiple exchanges to support higher level tools like placing trades/orders against that blockchain.

[btc_artist]

There's no reason there can't be more than one stock exchange.

Which is why I think a hybrid solution may be the best bet.

Have the stock ownership and transfers handled by a decentralized blockchain and then allow multiple exchanges to support higher level tools like placing trades/orders against that blockchain.

I agree.

But even if GLBSE wants nothing to do with a blockchain based solution, they could keep running.

[DeathAndTaxes]

There's no reason there can't be more than one stock exchange.

Which is why I think a hybrid solution may be the best bet.

Have the stock ownership and transfers handled by a decentralized blockchain and then allow multiple exchanges to support higher level tools like placing trades/orders against that blockchain.

I agree.

But even if GLBSE wants nothing to do with a blockchain based solution, they could keep running.

True but I hope they would see the advantage (maybe not personal advantage) of having the actual "shares" exist outside of GLBSE.  Imagine if for whatever reason GLBSE goes down forever tomorrow.  The chaos and yet more negative PR that brings to Bitcoin.

[btc_artist]

There's no reason there can't be more than one stock exchange.

Which is why I think a hybrid solution may be the best bet.

Have the stock ownership and transfers handled by a decentralized blockchain and then allow multiple exchanges to support higher level tools like placing trades/orders against that blockchain.

I agree.

But even if GLBSE wants nothing to do with a blockchain based solution, they could keep running.

True but I hope they would see the advantage (maybe not personal advantage) of having the actual "shares" exist outside of GLBSE.  Imagine if for whatever reason GLBSE goes down forever tomorrow.  The chaos and yet more negative PR that brings to Bitcoin.

I agree 100%.

[ripper234]

Just wanted to point everyone watching this thread to this new thread by btc_artist about DBSE

[Nefario]

There's no reason there can't be more than one stock exchange.

Which is why I think a hybrid solution may be the best bet.

Have the stock ownership and transfers handled by a decentralized blockchain and then allow multiple exchanges to support higher level tools like placing trades/orders against that blockchain.

I agree.

But even if GLBSE wants nothing to do with a blockchain based solution, they could keep running.

True but I hope they would see the advantage (maybe not personal advantage) of having the actual "shares" exist outside of GLBSE.  Imagine if for whatever reason GLBSE goes down forever tomorrow.  The chaos and yet more negative PR that brings to Bitcoin.

Well technically AFAIK when shares are issued and traded on the markets more than one organisation knows who owns the shares, including the company for who they are issued from. Are paper shares still given out anymore?

[markm]

If merged mining can work with arbitrary numbers of blockchains, maybe shares could be issued simply by starting a new chain that issues that many units in its genesis block? The issuing ocmpany would own those, and ownership can spread from there by normal operations of changing ownership on blockchains.

There are already supposedly methods one can use to lock in change of ownership on one chain so a change of ownership of something on another chain has to happen too.

Basically each type of share would be another alt currency, like the GMC (General Mining Corp) and GRF (General Retirement Funds) chains.

-MarkM-

[jtimon]

@markm
But what's wrong with having various types shares within the same chain?

Remember that even with merged mining miners need an incentive to secure the chain. Are the fees for trading a single company going to be enough to secure a chain?

[markm]

Open Transactions uses cryptography to sign everything, maybe just publishing transactions can suffice?

Anyone interested could keep histories, even though Open Transactions uses triple-signed receipts to avoid actually needing to keep transaction histories. (The receipt, signed by the two parties to the transaction and the server that executed it, serves as "proof" both parties and the notary server all agreed as to the balance, making the history of how that balance was arrived at moot/irrelevant.)

Since we depend on the issuer anyway - the company of which shares are issued - they could simply run a server themselves. They need not though, as any other server can allow them to do an issue. Either way there is their signed contract describing the issue, and triple-signed transactions whenever any units of the issue change ownership.

Ultimately we probably don't really care much how many people or entities who can do nothing in the way of actually grabbing for us some portion of the real assets of a company we have shares in think we have such shares. It is those who can actually give us something real when the company dies that need to know how much of the assets should be shipped to us or sold for something that is shipped/transferred to us.

But if all the transactions get published, so that the issuer and anyone else can see a server never issued any more than the actual issuer of the contract ordered them to, does it really matter what kind of storage or database various observers choose to use to store that info?

If a company's only assets are to be p2p assets though, such as holdings of various alt coins for example, then possibly it might make sense to try to run the whole company as a p2p application. As soon as you want it to hold non p2p assets though, there is centralisation, in that the value of each such asset is probably centralised at the location of that asset and/or the legal/de-facto ownership/control of that asset.

("Smart assets" can help with that though so maybe we should work on developing those while at it.)

-MarkM-

[markm]

If merged mining can work with arbitrary numbers of blockchains, maybe shares could be issued simply by starting a new chain that issues that many units in its genesis block? The issuing ocmpany would own those, and ownership can spread from there by normal operatinos of changing ownership on blockchains.

There are already supposedly methods one can use to lock in change of ownership on one chain so a change of ownership of something on another chain has to happen too.

Basically each type of share would be another alt currency, like the GMC (General Mining Corp) and GRF (General Retirement Funds) chains.

-MarkM-

I thought about this, but could this be susceptible to the 51% attack?

So far, yes. That is why GMC and GRF chains have been kept private, and now actually might even get abstracted away completely in favour of simply trading in Open Transactions some tokens representing such coins. The actual chain with the actual coins is starting to look like it can simply go away as not really being needed. Though if at some time such a chain does seem useful it can be created from scratch and the tokens in the Open Transactions system can be used to determine how many initial "coins" to issue initially to who.

Similarly with a number of other private blockchains; it is starting to look as if the actual chains can be ignored in favour of working directly with the Open Transactions tokens representing the values that would/could be represented in such chains.

If/when actual chains can be deployed without worry about 51% attacks, fine, that can be done. Meanwhile trading can proceed without waiting for the actual deployment of such a chain and such chains might even end up never actually being needed.

-MarkM-

[muyuu]

Hey buddies.

So my 2c is that this is not a technical problem.

Bitcoin is already the decentralised medium of exchange and works great for that.

The reason GLBSE has been kind of crappy is not because it's centralised but because it's just crappy. I made a much more in dept post as to where I went wrong and where I'll be improving GLBSE here.
https://bitcointalk.org/index.php?topic=52310.msg628049#msg628049

Bitcoin has come along and is the first working non-centralised solution to an already existing very centralised world (banking and money).

Markets ALL markets are by their very definition centralised, that's the whole point, everyone comes to the one place to trade, that way they will find a buy for their sell and vice versa.

The problems here are to do with size, as in the small size of the market (small markets don't thrive).
Enforcing contracts
Anonymity

They are not technical, at least not yet.

Nefario

Heh. Well, live and learn.

[jtimon]

Maybe some of the thread subscribers are interested in pybond, a project by jgarzik that can solve this problem. Some related threads:

https://bitcointalk.org/index.php?topic=92421.0
https://bitcointalk.org/index.php?topic=112007.0
https://bitcointalk.org/index.php?topic=106373.0
https://bitcointalk.org/index.php?topic=106449.0
https://bitcointalk.org/index.php?topic=114571.0

[markm]

Re Meni's concerns about commited offers and cancelling offers:

Cannot people who want to offer some bitcoins for, say, some litecoins use the cross-chain atomic trades system?

Or, if offering some blue satoshis for some red satoshis, use the in-chain atomic colour-trade system?

Both systems take a few steps, don't they? Until all the steps are complete the transaction the two (or potentially more?) parties are constructing cannot be mined into a block, right?

So one could have countless numbers of partially constructed atomic trades in progress at any particular moment, constantly optimising/prioritising how many steps toward completion/finalisation against how good a deal any particular atomic trade currently being constructed would be compared to others if it does end up being the one that gets finalised.

Cancelling is, surely, simply a matter of spending out from under it the outputs it is secifying as its inputs, so if someone does not finalise an atomic trade sooner than you spend out from under it the coins it planned to use as its input tough, it is invalidated thus in effect cancelled?

-MarkM-

[Meni Rosenfeld]

Re Meni's concerns about commited offers and cancelling offers:

Cannot people who want to offer some bitcoins for, say, some litecoins use the cross-chain atomic trades system?

Or, if offering some blue satoshis for some red satoshis, use the in-chain atomic colour-trade system?

Both systems take a few steps, don't they? Until all the steps are complete the transaction the two (or potentially more?) parties are constructing cannot be mined into a block, right?

So one could have countless numbers of partially constructed atomic trades in progress at any particular moment, constantly optimising/prioritising how many steps toward completion/finalisation against how good a deal any particular atomic trade currently being constructed would be compared to others if it does end up being the one that gets finalised.

Cancelling is, surely, simply a matter of spending out from under it the outputs it is secifying as its inputs, so if someone does not finalise an atomic trade sooner than you spend out from under it the coins it planned to use as its input tough, it is invalidated thus in effect cancelled?

-MarkM-

Well it's hard to discuss this when we're not sure what the cross-chain system will look like.

But at the basic level, once everything is said and done every party will need to sign the constructed transaction. If he can back away on this last step all the preceding steps aren't committing.

This can be resolved if a signature is provided beforehand with a script saying this is valid as long as anyone pays quantity X of Y to Z. Maybe it's possible but it will surely be complicated (even if it's within the same blockchain, the script will have to recognize the color of coins).

[jtimon]

Cancelling is, surely, simply a matter of spending out from under it the outputs it is secifying as its inputs, so if someone does not finalise an atomic trade sooner than you spend out from under it the coins it planned to use as its input tough, it is invalidated thus in effect cancelled?

That should be enough IMO.

Well it's hard to discuss this when we're not sure what the cross-chain system will look like.

We mean this, not currently possible (nLockTime and tx replacements are disabled).

But at the basic level, once everything is said and done every party will need to sign the constructed transaction. If he can back away on this last step all the preceding steps aren't committing.

This can be resolved if a signature is provided beforehand with a script saying this is valid as long as anyone pays quantity X of Y to Z. Maybe it's possible but it will surely be complicated (even if it's within the same blockchain, the script will have to recognize the color of coins).

For the all assets in one chain case (possible right now, that's what pybond tries to do), the transaction must be constructed carefully to take the colours into account, but nothing special is required for the signatures, just the default SIGHASH_ALL and all parties involved signing. If anyone moves its coins before the transaction gets into the block it is canceled no matter if all the signatures have been added or not.

[Meni Rosenfeld]

For the all assets in one chain case (possible right now, that's what pybond tries to do), the transaction must be constructed carefully to take the colours into account, but nothing special is required for the signatures, just the default SIGHASH_ALL and all parties involved signing. If anyone moves its coins before the transaction gets into the block it is canceled no matter if all the signatures have been added or not.

I'm not sure we're talking about the same thing. As I explained several times, making sure nobody runs away with the money is not a problem, that's easily solvable with both transfers being in the same transaction. I'm talking about manipulating the price by broadcasting orders you have no intention to execute, backing out when a 2nd party wants to take you on your offer.