7 ms·

What's the point of a private blockchain? Who are you competing against with your hash rate? And if you are keeping it private to stay in control, why choose a

by 0x0 10y ago

What's the point of a private blockchain? Who are you competing against with your hash rate? And if you are keeping it private to stay in control, why choose a technology designed to not let you stay in control? I've heard buzz about "enterprise blockchains" several times, but I still don't understand what this brings that an authenticated REST service on top of an SQL database doesn't.

rabboRubble 10y ago
In finance, successful two party private systems / standards are often turned into industry standards. Think Swift, FIX, etc. In the case of FIX, the reason it was widely adopted was that a buy-side client (Fidelity) demanded their brokers (developed with Goldman, then taken up by others) adopt the standard. I understand why Microsoft would want to support this. I can understand why BoAML would want this. Where I'm failing to see how this initiative will succeed is there seem to be missing a buy side participant to demand implementation. Perhaps BoAML has some sort of leverage with other market participants to get them on board? Recording transactions from one arm of BoAML with another arm of BoAML doesn't make sense to me. There are already internal systems to handle transaction recording.
conistonwater 10y ago
Fees, dude, fees! (More directly, it's not easy to sell nor charge a lot of money for just a database, so this is a genuine advance in sales technique.)
- wmf 10y ago
  I thought it's the opposite: banks want to stop paying fees to trusted third parties who run central settlement databases.
  - conistonwater 10y ago
    I might be confused: I thought the only reason those third parties were trusted in the first place was that they were co-owned by all the banks? My point was, you replace one cooperative system with another functionally equivalent one, and nothing much changes except that there are fees to be collected in the process. A centralized settlement system is like a commodity, it's standardized and cheap to run, but a blockchain can be sold. There's also an unjustified assumption that the new system wouldn't have banks paying fees, because blockchain. There's a long series of posts on Alphaville about just that. All those trust and legal issue don't just disappear.
    - AtheistOfFail 10y ago
      Think about this, burning power in every office while gaining a 100x redundancy factor may be worth it. The cost of going from 99% to 99.999% can be a 100000x increase in the software space. For reference Bitcoin has a 99.99% uptime per http://www.bitcoinuptime.com/ http://www.bitcoinuptime.com/
  - dragonwriter 10y ago
    So, instead they are going to pay fees to third parties who provide cloud-based "blockchain-as-a-service" which functions as a centralized (on Azure) settlement database?
milcron 10y ago
It's possible to design a blockchains that don't require brute-force computation. Bitcoin uses a "proof of work" algorithm, but Peercoin (for instance) uses "proof of stake" which is a lot less CPU-intensive.
- brighton36 10y ago
  And peercoin stake grinded to oblivion. Andrew Poelstra writes extensively about the problems with non-work systems. (which are ancient technology)
  - milcron 10y ago
    Are there any active attempts at finding alternatives to proof-of-work?
    - jude- 10y ago
      The fundamental problem seems to be that, when positioned against the same threat model as Bitcoin, all the PoW alternatives reduce to an obscured form of PoW. Put another way, a PoW alternative proposal must be designed such that participants can't increase profits expending more energy (for example, the fact that stake-grinding is profitable shows that Peercoin's PoS is really PoW-by-obscurity). As far as I know, this is still an open problem. There's an interesting primer on the subject here: http://www.truthcoin.info/blog/pow-cheapest/ http://www.truthcoin.info/blog/pow-cheapest/
    - aeijdenberg 10y ago
      If you can accept centralization (which for many non-crypto-currency use-cases actually makes sense, since it mirrors the way the real world works) - you don't actually need proof of work. You can still get the benefits of verifiability, append-only, detect misbehavior etc without it. See for example Certificate Transparency (RFC6962) for a standard that implements those properties using similar Merkle Tree constructions. Disclaimer, I have a startup that generalizes the same Verifiable Logs (and also allows for Verifiable Maps): https://www.continusec.com/ https://www.continusec.com/
pjc50 10y ago
Presumably you'd ditch the proof-of-waste system entirely and just have a short whitelist of keys allowed to post to it. It sounds like a means of getting a distributed database that actually works, with no central admin. That's no small thing.
- 0x0 10y ago
  What, like git?
  - oconnore 10y ago
    Probably like git but with every commit signed by n+1 / 2n participants.
    - therein 10y ago
      So git with a few pre-commit hooks?
      - verroq 10y ago
        And Dropbox is literally rsync am I right?
  - pjc50 10y ago
    Git doesn't currently handle "accounts", it handles lines of text in diff form. You can't currently use git to pay for things, though no doubt someone will add this feature.
- brighton36 10y ago
  Proof of Work is the process by which registered value (an electric bill) is converted to fungible value. It is nearly 100% efficient (depending on how you quantify waste). If you don't like the externality costs of mining - I'm with you. But that's the government's job to manage, and seemingly, they're ok with that. They even subsidize the network by requiring that some goods only be purchased via blockchain.
cloudjacker 10y ago
well this is built on the Ethereum network, according to a tweet I saw in this thread, and would therefore be a private contract. So the private data is settled on a blockchain that is secured by proof of work now, and proof of stake later.
ThomPete 10y ago
One of them I think is to establish the standard for true international wireless transfers protocol since they don't really exist today[1]. This makes a huge difference when it comes to something like remittance and would allow for a whole new kind of international companies and collaboration to be built. [1] http://paymentsviews.com/2014/05/15/there-is-no-such-thing-as-an-international-wire/ http://paymentsviews.com/2014/05/15/there-is-no-such-thing-a...
brighton36 10y ago
it's just a buzzword. These are simply bad message passing systems that will never come to fruition as a blockchain. R3 spearheaded this effort and already dropped the unneccessary 'blocks' from their project. I'm unclear what their pitch is now, other than sharing your data with your competitors. A stupid goal that could be better performed with http..
- mistermumble 10y ago
  R3 is now emphasizing the smart contract aspect, and is building their system on Java Virtual Machine (JVM). If blockchain technology is all about adding trust in a trustless environment, what are the trust implications of using Oracle JVM?
  - kpil 10y ago
    That it's fine? The problem have always been running others untrusted code in your own jvm.
- eternalban 10y ago
  Was interviewing with a blockchain company yesterday and this was my question to them. There is a strong buzz -- imo not very informed buzz -- about this approach in the enterprise and after this cycle of failed pilot projects the buzz should/will die. Blockchains waste a substantial amount of computing resources. All that heat is wasted energy if the technology is used where the unique featureset of a blockchain are not important. And in the enterprise, that unique featureset is typically not really that important. B2B is a much better enterprise space for blockchains since the name of the blockchain game is 'distributed computing at the edge'.
ledgerdev 10y ago
The value prop for private or consortium chains is significantly different than public blockchains. The first point is that it's a shared database that allows organizations to co-operate, with a common view of the world meaning fewer integration points which are exponential as the number of participants grows. The second point is that none of the individual parties has monopoly control of the "central" database meaning they can't extract excessive rent from the participants and exert political/corrupt influence on the network. This is my favorite video explaining this and blockchains in general, and this aspect specifically which begins at 23:45 in. The sound is terrible, but definitely worth listening to. https://vimeo.com/153600491 https://vimeo.com/153600491
- runeks 10y ago
  The thing is, none of the things you mention require proof-of-work, which is really the only thing that makes a Blockchain a Blockchain, rather than just a weird append-only database which allocates tokens to accounts (scripts). Only by proof-of-work do these tokens become meaningful, because it allows the tokens to be moved around by the account-holders without a central party. There is simply no need for a private consortium to waste energy on proof-of-work. The features you mention can be implemented without any need for proof-of-work. The purpose of the organized waste that is proof-of-work is to achieve trustlessness, simply by requiring a huge expansion of energy if you wish to rewrite history. In Bitcoin, the maintainers of the database (the miners) are separate from the users, and proof-of-work prevents the Blockchain/database maintainers (miners) from rewriting history. But if the owners and users of the database exist within the same organization, it would be trivial to verify that nothing has been rewritten simply by every party keeping a copy.

by 0x0 10y ago

rabboRubble 10y ago

In finance, successful two party private systems / standards are often turned into industry standards. Think Swift, FIX, etc. In the case of FIX, the reason it was widely adopted was that a buy-side client (Fidelity) demanded their brokers (developed with Goldman, then taken up by others) adopt the standard. I understand why Microsoft would want to support this. I can understand why BoAML would want this. Where I'm failing to see how this initiative will succeed is there seem to be missing a buy side participant to demand implementation. Perhaps BoAML has some sort of leverage with other market participants to get them on board? Recording transactions from one arm of BoAML with another arm of BoAML doesn't make sense to me. There are already internal systems to handle transaction recording.

conistonwater 10y ago

Fees, dude, fees! (More directly, it's not easy to sell nor charge a lot of money for just a database, so this is a genuine advance in sales technique.)

wmf 10y ago
I thought it's the opposite: banks want to stop paying fees to trusted third parties who run central settlement databases.
- conistonwater 10y ago
  I might be confused: I thought the only reason those third parties were trusted in the first place was that they were co-owned by all the banks? My point was, you replace one cooperative system with another functionally equivalent one, and nothing much changes except that there are fees to be collected in the process. A centralized settlement system is like a commodity, it's standardized and cheap to run, but a blockchain can be sold. There's also an unjustified assumption that the new system wouldn't have banks paying fees, because blockchain. There's a long series of posts on Alphaville about just that. All those trust and legal issue don't just disappear.
  - AtheistOfFail 10y ago
    Think about this, burning power in every office while gaining a 100x redundancy factor may be worth it. The cost of going from 99% to 99.999% can be a 100000x increase in the software space. For reference Bitcoin has a 99.99% uptime per http://www.bitcoinuptime.com/ http://www.bitcoinuptime.com/
- dragonwriter 10y ago
  So, instead they are going to pay fees to third parties who provide cloud-based "blockchain-as-a-service" which functions as a centralized (on Azure) settlement database?

milcron 10y ago

It's possible to design a blockchains that don't require brute-force computation. Bitcoin uses a "proof of work" algorithm, but Peercoin (for instance) uses "proof of stake" which is a lot less CPU-intensive.

brighton36 10y ago
And peercoin stake grinded to oblivion. Andrew Poelstra writes extensively about the problems with non-work systems. (which are ancient technology)
- milcron 10y ago
  Are there any active attempts at finding alternatives to proof-of-work?
  - jude- 10y ago
    The fundamental problem seems to be that, when positioned against the same threat model as Bitcoin, all the PoW alternatives reduce to an obscured form of PoW. Put another way, a PoW alternative proposal must be designed such that participants can't increase profits expending more energy (for example, the fact that stake-grinding is profitable shows that Peercoin's PoS is really PoW-by-obscurity). As far as I know, this is still an open problem. There's an interesting primer on the subject here: http://www.truthcoin.info/blog/pow-cheapest/ http://www.truthcoin.info/blog/pow-cheapest/
  - aeijdenberg 10y ago
    If you can accept centralization (which for many non-crypto-currency use-cases actually makes sense, since it mirrors the way the real world works) - you don't actually need proof of work. You can still get the benefits of verifiability, append-only, detect misbehavior etc without it. See for example Certificate Transparency (RFC6962) for a standard that implements those properties using similar Merkle Tree constructions. Disclaimer, I have a startup that generalizes the same Verifiable Logs (and also allows for Verifiable Maps): https://www.continusec.com/ https://www.continusec.com/

pjc50 10y ago

Presumably you'd ditch the proof-of-waste system entirely and just have a short whitelist of keys allowed to post to it. It sounds like a means of getting a distributed database that actually works, with no central admin. That's no small thing.

0x0 10y ago
What, like git?
- oconnore 10y ago
  Probably like git but with every commit signed by n+1 / 2n participants.
  - therein 10y ago
    So git with a few pre-commit hooks?
    - verroq 10y ago
      And Dropbox is literally rsync am I right?
- pjc50 10y ago
  Git doesn't currently handle "accounts", it handles lines of text in diff form. You can't currently use git to pay for things, though no doubt someone will add this feature.
brighton36 10y ago
Proof of Work is the process by which registered value (an electric bill) is converted to fungible value. It is nearly 100% efficient (depending on how you quantify waste). If you don't like the externality costs of mining - I'm with you. But that's the government's job to manage, and seemingly, they're ok with that. They even subsidize the network by requiring that some goods only be purchased via blockchain.

cloudjacker 10y ago

well this is built on the Ethereum network, according to a tweet I saw in this thread, and would therefore be a private contract. So the private data is settled on a blockchain that is secured by proof of work now, and proof of stake later.

ThomPete 10y ago

One of them I think is to establish the standard for true international wireless transfers protocol since they don't really exist today[1]. This makes a huge difference when it comes to something like remittance and would allow for a whole new kind of international companies and collaboration to be built. [1] http://paymentsviews.com/2014/05/15/there-is-no-such-thing-as-an-international-wire/ http://paymentsviews.com/2014/05/15/there-is-no-such-thing-a...

brighton36 10y ago

it's just a buzzword. These are simply bad message passing systems that will never come to fruition as a blockchain. R3 spearheaded this effort and already dropped the unneccessary 'blocks' from their project. I'm unclear what their pitch is now, other than sharing your data with your competitors. A stupid goal that could be better performed with http..

mistermumble 10y ago
R3 is now emphasizing the smart contract aspect, and is building their system on Java Virtual Machine (JVM). If blockchain technology is all about adding trust in a trustless environment, what are the trust implications of using Oracle JVM?
- kpil 10y ago
  That it's fine? The problem have always been running others untrusted code in your own jvm.
eternalban 10y ago
Was interviewing with a blockchain company yesterday and this was my question to them. There is a strong buzz -- imo not very informed buzz -- about this approach in the enterprise and after this cycle of failed pilot projects the buzz should/will die. Blockchains waste a substantial amount of computing resources. All that heat is wasted energy if the technology is used where the unique featureset of a blockchain are not important. And in the enterprise, that unique featureset is typically not really that important. B2B is a much better enterprise space for blockchains since the name of the blockchain game is 'distributed computing at the edge'.

ledgerdev 10y ago

The value prop for private or consortium chains is significantly different than public blockchains. The first point is that it's a shared database that allows organizations to co-operate, with a common view of the world meaning fewer integration points which are exponential as the number of participants grows. The second point is that none of the individual parties has monopoly control of the "central" database meaning they can't extract excessive rent from the participants and exert political/corrupt influence on the network. This is my favorite video explaining this and blockchains in general, and this aspect specifically which begins at 23:45 in. The sound is terrible, but definitely worth listening to. https://vimeo.com/153600491 https://vimeo.com/153600491

runeks 10y ago
The thing is, none of the things you mention require proof-of-work, which is really the only thing that makes a Blockchain a Blockchain, rather than just a weird append-only database which allocates tokens to accounts (scripts). Only by proof-of-work do these tokens become meaningful, because it allows the tokens to be moved around by the account-holders without a central party. There is simply no need for a private consortium to waste energy on proof-of-work. The features you mention can be implemented without any need for proof-of-work. The purpose of the organized waste that is proof-of-work is to achieve trustlessness, simply by requiring a huge expansion of energy if you wish to rewrite history. In Bitcoin, the maintainers of the database (the miners) are separate from the users, and proof-of-work prevents the Blockchain/database maintainers (miners) from rewriting history. But if the owners and users of the database exist within the same organization, it would be trivial to verify that nothing has been rewritten simply by every party keeping a copy.