Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Facebook plans cryptocurrency debut

techcrunch.com

Post reply on HN
Print view View original

41–50 of 345 posts

Re: Facebook plans cryptocurrency debut

#41

Post by bigiain »

mbrock wrote: post #36
wz1000 wrote: post #29

Earlier quoted context omitted.

Every git repository counts as a blockchain by that definition.

Indeed you could even think of them as embodying the "proof of work" principle in that you naturally look for the branch with the most work done to find the true Schelling point fork for your own contributions. (By the way, I often see these two points in the same thread: "blockchains should be replaced with centralized SQL databases" and "Git repositories are just like blockchains" which makes me wonder if Git repos…

Aka Perforce...

Re: Facebook plans cryptocurrency debut

#42

Post by mbrock »

lowracle wrote: post #32
kolinko wrote: post #18

A few points: - some pegged currencies are still decentralised (e.g. MakerDAO) - they rely solely on smart contracts - even with a centralised peg, you have to trust an organisation to keep the collateral, but the contract can state that anyone can move the funds around, so the managing organisation cannot just remove funds at will Also, compared to a centralised database, you have an interoperability aspect - if it'…

MakerDAO is not decentralised, simply because the ETH/USD price feed isn't decentralised. If the servers providing the feed get hacked, say goodbye to your ETH collateral.

Stakeholder vote elects a smart contract that takes the median of dozens of price feeds controlled by various independent parties, so it's decentralized to some extent, and decentralization is not a clear-cut binary attribute.

Re: Facebook plans cryptocurrency debut

#43

Post by rusticpenn »

alehul wrote: post #31
CaptainZapp wrote: post #27

Earlier quoted context omitted.

Have you transferred money across borders to an individual lately? Actually yes. From Switzerland to my brother in Belgium. It was transferred on the same business day at no expense to him or me. Look up SEPA[1] for details. Just because the American banking system is the complete shits this does not necessarily apply to the rest of the world. [1] https://en.wikipedia.org/wiki/Single_Euro_Payments_Area

Across very cooperative borders in wealthy Europe, sure. Do you think it's as easy for those in India, China, or Kenya, for instance?

India has paytm, There are providers like worldremit who transfer money in a few hours. ( transfered recently from Europe to India)

Re: Facebook plans cryptocurrency debut

#44

Post by mbrock »

wz1000 wrote: post #40
pauletienney wrote: post #34

Earlier quoted context omitted.

except for "append-only" and "verifiable records"

Git is "append-only". Just like a normal blockchain, you can't modify an older commit without rewriting all the commits since then. Git also supports signed commits.

So Git could be a somewhat reasonable backend for a cryptocurrency... if only its hash function weren't broken.

Re: Facebook plans cryptocurrency debut

#45

Post by momokoko »

wz1000 wrote: post #29
mbrock wrote: post #23

Facebook's blockchain will most likely be a consortium of corporate validators. In this case you could say that the blockchain is "just" a normal distributed append-only hash-linked ledger with public verifiable records, but nowadays the more concise term for that is "blockchain."

Every git repository counts as a blockchain by that definition.

Merkle Tree.

Re: Facebook plans cryptocurrency debut

#46

Post by rubber_duck »

mbrock wrote: post #36
wz1000 wrote: post #29

Earlier quoted context omitted.

Every git repository counts as a blockchain by that definition.

Indeed you could even think of them as embodying the "proof of work" principle in that you naturally look for the branch with the most work done to find the true Schelling point fork for your own contributions. (By the way, I often see these two points in the same thread: "blockchains should be replaced with centralized SQL databases" and "Git repositories are just like blockchains" which makes me wonder if Git repos…

Well git model has a lot of issues with scaling repo size - git extensions like GVFS and LFS are used to remedy the problem but they break the decentralized part so you have this hybrid solutions.

Re: Facebook plans cryptocurrency debut

#47

Post by wcoenen »

CaptainZapp wrote: post #27
A2017U1 wrote: post #13

Earlier quoted context omitted.

Have you transferred money across borders to an individual lately? As much as I want to criticize this FB is probably in an excellent position to navigate the kyc minefields.

Have you transferred money across borders to an individual lately? Actually yes. From Switzerland to my brother in Belgium. It was transferred on the same business day at no expense to him or me. Look up SEPA[1] for details. Just because the American banking system is the complete shits this does not necessarily apply to the rest of the world. [1] https://en.wikipedia.org/wiki/Single_Euro_Payments_Area

SEPA is okay, but it's typically not "same business day". Just this week I initiated a transfer on Tuesday and it only arrived Thursday morning. This was between two Belgian accounts.

Re: Facebook plans cryptocurrency debut

#48

Post by m-i-l »

wz1000 wrote: post #6

Can anyone explain what the point of a pegged cryptocurrency is? Since you are dependent on the issuing entity to honour the stated peg, why not just have your balance be an entry in a regular database maintained by that entity? What exactly does the blockchain gain you here? One thing it does _not_ get you is censorship resistance. If the central authority wants to take away coins from a wallet, all it needs to do i…

> Can anyone explain what the point of a pegged cryptocurrency is? Since you are dependent on the issuing entity to honour the stated peg

In this case there wouldn't be a single issuing entity: "the cryptocurrency would indeed be pegged to a basket of currencies". I had written about the idea of a stablecoin tied to a basket of currencies, e.g. J M Keynes' Bancor and the Word Currency Unit, a few years back.

> why not just have your balance be an entry in a regular database maintained by that entity? What exactly does the blockchain gain you here?

If it is controlled by a centralised organisation there would appear to be little point of "blockchain", but maybe as the article says "By introducing a level of decentralization to the governance of the project, Facebook may be able to avoid regulation related to it holding too much power over a global currency".

Re: Facebook plans cryptocurrency debut

#49

Post by vcavallo »

mihaifm wrote: post #9

If you're interested in conspiracy theories, this may be a way for Facebook to circumvent GDPR, since financial data needs to be kept for at least 10 years. Once you make any trade using this currency, you won't be able to request deletion of personal data, since it's required to be kept as financial information.

why couldn’t they delete everything but your trade data?

Re: Facebook plans cryptocurrency debut

#50

Post by wz1000 »

mbrock wrote: post #36
wz1000 wrote: post #29

Earlier quoted context omitted.

Every git repository counts as a blockchain by that definition.

Indeed you could even think of them as embodying the "proof of work" principle in that you naturally look for the branch with the most work done to find the true Schelling point fork for your own contributions. (By the way, I often see these two points in the same thread: "blockchains should be replaced with centralized SQL databases" and "Git repositories are just like blockchains" which makes me wonder if Git repos…

> blockchains should be replaced with centralized SQL databases

Blockchains whose tokens are subject to the authority of a central issuer anyway should be replaced with centralized SQL databases. Bitcoin/namecoin/etc. can't be replaced with centralized databases. FBCoin can, as can any blockchain that is supposed to track meatspace resources(since physical resources are subject to the authority of the local government/men with guns anyway).

> Git repositories are just like blockchains

Nope, I don't consider this to be true. Blockchains are supposed to have decentralized consensus and allow arbitrary nodes to propose changes.

Post reply on HN