Topic: colored bitcoin tech discussion (Read 13114 times)

Well, what kind of reply do you expect?

It looks like the idea behind Confidence Chains is "Let's just trust some guys to secure transaction processing", and it just isn't something I'm interested working on, sorry.

(Besides that, I believe there are major flaws in Confidence Chains design, but I'm not going to discuss them in this thread as this would be off-topic; and we already discussed them in bitcoinX mailing list anyway.)

---

Another reason for me to work on colored coins specifically is that I don't have enough confidence to work on a completely new cryptocurrency. The good thing about colored coins is that they inherit almost all properties from Bitcoin itself. I like the simplicity of design.

As for capacity issues, I already explained that I do not find them important. When Satoshi released Bitcoin people complained that it won't scale, but look at where Bitcoin is now.

I prefer to do the simple thing first. Is it useful? Yes, it is.

It doesn't solve everything, of course. But I prefer multi-layer design to a monolithic thing which would try to do everything at once. (I don't think that a decentralized replacement for NYSE is even possible, by the way.)

---

I think it's fairly obvious that you're trying to criticize colored coins to promote your Confidence Chains. Well, do you know why people ignore you? Is it some kind of a global conspiracy? No, it's just an obviously bad design.

But the general idea of trust-based consensus is sensible; and if you're interested in things like that, you might be able to cooperate with others and make something good. But to do that you need to be ready to ditch your own ideas and listen to others. I hope you realize that spending all efforts to push the first thing which came to your mind isn't exactly a healthy thing to do sooner rather than later.

(FYI before I started working on colored coins, I was working on an  alt-coin design and on distributed Ripple analog based on merged mining (Ripple.com wasn't public back then). I still believe I had some good ideas, but colored coins just had much higher chances of success, so that's what I focused on.)

Why would you want to:

 1) implement a complex technology to retrofit to existing Bitcoin

 2) that may have catastrophic capacity issues and introduce a whole new tier of policing problems to Bitcoin

 3) that requires a THIRD or even a FOURTH technology(some of which themselves are unproven or even unspecified) in order to achieve auxiliary functions

Just to do what Confidence Chains does simply, elegantly, efficiently and natively?

What we have now in NGCCC is exactly the same thing which I implemented ago (ArmoryX), it is 100% compatible.

We just got better terminology, better understanding how to handle various corner cases, future improvements, better wallet design etc.


Well, the thing is, I'm simply implementing an open source tool for asset tracking on top of Bitcoin blockchain.

I think it's obvious that there are inherent limitations. Say, if BItcoin block size is limited to 1 MB, Bitcoin cannot process more than 7 transactions per second on average. And as each colored coin transaction is a Bitcoin transaction, we cannot have more than 7 colored coin transactions per second either.

So, obviously, it cannot replace NYSE :-), but it doesn't mean that it isn't useful.

For example, existing Bitcoin companies simply do not have trade volumes which would get anywhere close to the limits imposed by the blockchain, so it's quite possible to represent their shares with colored coins and do this decentralized trading thing.

For example, one of the more popular stocks -- AM100 shares on HavelockInvestments -- had just 20 trades in last 6 hours.

So that what I'm doing now: a tool which can be useful in some contexts.

---

Of course, sometimes people ask: what if we need to handle more than 7 transactions per second? Is it possible to do that?

Well, it is possible to do transactions off-chain (i.e. without publishing real Bitcoin transactions), but that requires some trade-offs in decentralization and security.

For example, it is possible to combine colored coins with Open Transactions servers: colored coins will be used to move assets between different servers and long-term wallets, while the bulk of trading will be done within servers.

Sure, it isn't as bulletproof as native colored coins trade, but Open Transactions will be a massive improvement compared to 100% proprietary and opaque platforms like BitFunder, MtGox etc.

In any case, it is important to understand that I'm not trying to sell a complete solution, I'm just developing open source software. It's up to people to decide how to use this software (and whether to use it).

the implementation technique has changed quite a few times.  The main concern with Color Coins is performance.  Think NumberTransactions ^ NumberCoinColors.  There haven't been any good studies on how this will pan out but that's the general estimate.  Seems there is a love/hate relationship between core Bitcoin devs and Color Coins, some seem to support it, others don't.  There was the COIN_DUST issue which factors into Color Coin's future.  Putting Color Coins on alt-chain is just pointless because if you have a new chain, you can redesign the tx format completely.  Color Coins main feature is backward compatibility with existing Bitcoin infrastructure.  

Alex has backpeddled a number of times on his statements and I don't spend much time reviewing his claims.  Keep in mind, money has been spent on this project so the investors do have an agenda here.

Actually we use the same thing, it's just more generalized now, and would allow different algorithms in future.


wat?

Thanks, but we are using a different model now: instead of per-task bounties, we pay according to number of hours spent on work.

As it turned out, it's much easier to hire freelancers than to do this bounty thing... It's now possible to get things done in matter of days, not weeks or months.

I don't know whether it would help much but CIYAM Open would be very happy to have the colored coins project listed (fee free for the life of the project).

PM me if you're interested to do that.

I see.  I was under the impression that the coloring algorithm in Meni Rosenfeld's paper "Overview of Colored Coins" was the one that's employed.  Thanks.

Here's an updated theoretic model: https://github.com/bitcoinx/colored-coin-tools/wiki/colored_coins_intro

Colorvalue is a deterministic function. For a given (connected) transaction output and a given color (e.g. "obc:cc8e64cef1a880f5132e73b5a1f52a72565c92afa8ec36c445c635fe37b372fd:0:263370") colorvalue is certain and is same for everyone.

There is, of course, another layer, which we call asset definition layer: people need to know that, say, 10000 atoms of "obc:cc8e64cef1a880f5132e73b5a1f52a72565c92afa8ec36c445c635fe37b372fd:0:263370" means 1 oz of gold backed by issuer ABC.

But this isn't a technical problem. User needs to decide whether he trusts issuer ABC. If he does, getting an authoritative list of colors from issuer ABC is pretty much trivial, for example, one can get them from ABC's web site, and they will be signed with his PGP key.

No, you are mistaken. There has to be agreement between two parties in an exchange as to what the color they are exchanging is. However coloring is deterministic, and does not affect bitcoin consensus in any way. So as long as you and the people you are interacting with are in agreement on the color definition, that is all that is required. You do not need global consensus on anything except the bitcoin block chain.

If I understand the coloring algorithm correctly, there has to be a global agreement as to what the colors are.  Otherwise if one party believed in yellow and another didn't (i.e., saw yellow coins as uncolored), they would disagree about how the colors propagated.  Do I have this right?

Investors are only interested in things which brings them profit.

It's very hard to extract profit out of fully open source project. It is theoretically possible, but not easy.

Mastercoin was able to get $600k for development through creation of a new currency.

But the whole point of colored coins is that they work directly with Bitcoin and are not limited in any way, so this model cannot be used.

We already have funding... 200 BTC was pledged by sponsors for a complete implementation:

1. 100 BTC for NGCCC: a client which can be used on desktops and servers
2. 100 BTC for WebcoinX, a web client

It is still hard to find developers!

100 BTC will buy you only 1-2 months of work of North American/European programmer...

But to have a good progress, we need developers who already have in-depth understanding of the Bitcoin protocol and everything which goes with it, and who are qualified to work on project like this. Of course, good devs are already working somewhere, and it's unlikely that they will consider leaving their current work to have 1 month gig. (Cf. J. R. Willet: he didn't abandon his work even though he got something like $600k to develop Mastercoin.)

So there are only people who can work in their free time, on weekends, etc. And, understandably, it doesn't result in fast development.

I think it's worth noting that we already have several implementations which are almost complete in terms of functionality, particularly:

1. ArmoryX: We got it working in January of 2013. It was able to do many things: issuing colored coins, installing "color definitions" from web, p2p trade. It had pretty much everything, except it lacked some checks to make it foolproof. We had to abandon it because Armory requires too much RAM and crashes too often, that basically makes it unusable... Also it is hard to get in sync with Armory mainline.

2. WebcoinX: We got a prototype working on testnet in May of 2013. It could do all things Armory could do, I think. But it is too slow to work on mainnet, and it wasn't secure... But wrapping it as a Chrome extension and fetching data in a different way isn't exactly rocket science. Still, we only got progress with bugfixing and GUI, it's still not ready for mainnet.

3. CCoinDemo is a really compact wallet... Basically, it is a thin layer on top of coloredcoinlib which uses web services and bitcoind RPC API. It's something like hundred lines of Python code.

So, there is no shortage of half-assed and mostly-working wallets, there is a shortage of people who will be committed to making a complete, reliable wallet.

I mean, I do what I can... But we could do it faster if we had more people....

(The last guy who fully understood how it works and was willing to write code disappeared in June... I'm talking about Karel Tuma who was leading WebcoinX development. There were do signs that he's going to abandon the project, so I assume that something have happened to him IRL. If not that, we'd probably already have WebcoinX reasonably complete.)

So now about hope: I hope to complete a basic core for NGCCC in a couple of days. I also found several developers interested in working on this. After we have a good core we can work incrementally.

Well a group of respected known faces of the project can create a multi-sig secure bitcoin address and then call for a kind of crowdfunding, it could be in IPO or any other form that suits the project characteristics better and also includes regulatory concern.
Mastercoin did such an IPO and got a huge funding capital, it was during August when BTCT and Bitfunder were working normally, a crowdfunding or kickstarter for coloredcoin should bring more attention in this desperate time.

Maybe they could list on an exchange

Why don't you try to crowdfund the project and then choose and hire some developers for a compact implementation? Due to recent events, timing is great.

Yes, there is no clear date.

Unfortunately there was a problem with recruiting developers, i.e. barely anybody was working on this for months, but there is some hope that it will change soon...

We'll probably have something available this month, but there are no guarantees...

What do you mean by "centralized exchanges"?

IBs (or intermediaries/proxies) will assist business with public offers, paperwork and whatnot. I think a good model for it is PicoStocks, and, maybe, MPEx. (But they are bundled with a trading platform.)

Rating agencies will help investors to find worthy securities: they can work with businesses and their representatives to perform audit, independent valuation analysis and so on. In the end, each rating agency will create a list of securities it believes are good.

If we assume that decentralized trading platforms exist, I see what role "centralized exchanges" might have. A security doesn't need to be listed in a specific place.

It can be listed on:

• rating agency's site (it is probably preferred, as rating agencies have strongest incentive to provide proper information)
• IB's site (they have an incentive to promote businesses they work with)
• own site of a company
• on bitcointalk

No matter where user gets information from, he can trade shares of that company using this decentralized trading software.


Who SEC will go after?

Are you aware of model which is used by PicoStocks?

They bought shares of CoinTerra, a private company registered in US, and now as a company registered on Marshall Islands they let users trade virtual shares based on real CoinTerra.

Do you think that SEC has an authority to close PicoStocks?

Proxies are not in any way unique to cryptocurrencies. E.g. Goldman Sachs created a fund as a proxy to Facebook shares when Facebook was private.

Then there is SecondMarket.com:


Look, I'm just making a tool, and I anticipate certain uses. But people will likely use it in many different ways, and it is hard to predict what will work and what won't.

In between, may I ask, if you have a kind of time table or a close-date for an alpha or beta release? Or not a clear implementation date yet?

IMHO businesses should not be directly managing their own listings. There are conflicts of interest and issues with ignorance to how to perform the role that are quite considerable. This inevitably puts IBs back into the picture, which begs for centralized exchanges to exist to separate powers. Either way, the SEC still has a focal point, and thus the decentralized system is no stronger than the current one in that regard.

This unicorn is a moving target. You can get close to it, but I'm not sure you can ever touch it.

I suppose what I'm saying results in suggesting that the solution is actually for a Bitcoin Stock Exchange to make the effort of trailblazing the path of legitimacy, and for Investment Banks to do the same. If decentralization motivations lead to giving those two powers better and more accountable tools, and lead to shareholders having a default contingency plan, great.

In that sense, do these decentralization efforts boil down to creating digitally verifiable stock certificates, and nothing more? If so, I think many people cheering for the effort have the wrong impression of what the movement will achieve...