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

Why aren’t distributed systems engineers working on blockchain technology?

eng.paxos.com

Post reply on HN
Print view View original

111–120 of 204 posts

Re: Why aren’t distributed systems engineers working on blockchain technology?

#111

Post by runeks »

devrandomguy wrote: post #44
wyager wrote: post #31

Earlier quoted context omitted.

Look up "Lightning Network". This is a solved problem. You can build on top of bitcoin to scale basically as large as you want, trustlessly. The software is being developed as we speak.

I sense incoming downvotes, it would be great if we could get a link to an existing rebuttal, or something. Stomping on someone's dream just so you don't have to risk disappointment holds us all back. What are the limits of the Lightning network?

The challenge for the Lightning Network is cheaply routing payments from consumers to merchants. LN is optimized for anyone-to-anyone payments, while in a real economy money moves from consumer to merchant to producer to worker (consumer). LN is not optimized for this circular pattern, where many consumers pay a single merchant.

Also, everything is done via two-way payment channels in LN, which means that you need to lock X BTC in a channel in order to send it, which becomes very capital intensive for nodes close to merchants who need to lock millions of dollars worth of bitcoins to cover merchants’ monthly turnover (this is in addition to the BTC locked up by other nodes who participate in the transfer but are not directly connected to merchants). If you want to send 0.1 BTC through 10 nodes, each of these 10 nodes must have locked up 0.1 BTC on the blockchain, and you’ll pay blockchain fees if just a single channel/node in your path doesn’t have at least as much as you need to send.

A proper payment network needs two different types of nodes: senders and receivers. Using a single type of node for both doesn’t make much sense, since merchants and consumers have completely different needs. Senders need to be able to send relatively small amounts relatively frequently (a consumer making a few purchases per day), while receivers (merchants) need to redeem relatively large amounts (e.g. one week’s worth of payments) relatively rarely for bitcoins (on the blockchain).

Re: Why aren’t distributed systems engineers working on blockchain technology?

#112

Post by vocatus_gate »

asah wrote: post #58

I had the same observation that the blockchain is "just" a distributed log database with horrific performance overhead... ...but it's not true that nobody's working on it: quietly, Serious Engineers(tm) working for blockchain companies and you can expect big improvements in the next few years. This rant is nearly isomorphic to the whining in the early 90s by the academics and SunRPC & CORBA fanboys vs HTTP and HTML,…

Can't remember how I stumbled across this article, but I thought it was a good summary of Bitcoin works in spite of some glaring flaws.

https://www.gwern.net/Bitcoin%20is%20Worse%20is%20Better

Re: Why aren’t distributed systems engineers working on blockchain technology?

#113

Post by svara »

sigstoat wrote: post #89
notzorbo3 wrote: post #74

Earlier quoted context omitted.

The article doesn't explain why a public distributed ledger such as the blockchain would make sense. A privately run ledger would probably also solve their problems, be cheaper and it would be a lot more secure against tampering.

which of the mutually untrustworthy parties will operate the ledger? customs agents don't trust shipping companies, shipping companies don't trust their customers, shipping companies might trust most customs agents in western countries but they operate everywhere, and there are, as i understand it, a number of other middle-men involved in the process, who probably aren't trusted much either. i don't see how a private…

So are you suggesting that shipping companies, their customers, customs and middlemen should be mining a cryptocurrency? Or that they should record their interactions on a blockchain running on a peer-to-peer network of random miners around the world?

Sorry, I just don't see that. There's laws, governments, clearing houses that have allowed untrustworthy parties to work together for... ever. And if they need a database, surely there's going to be a more pragmatic solution?

Re: Why aren’t distributed systems engineers working on blockchain technology?

#114

Post by redy »

ntrepid8 wrote: post #55
darawk wrote: post #43

Earlier quoted context omitted.

There are solutions to that though, like Proof of Stake. If distributed systems really cared about that, then that's exactly why they ought to be working on blockchains, to make them more efficient.

Doesn't blockchain verification depend on a certain amount of inefficiency? Isn't that what makes forgery difficult?

No serious blockchain relies on proof-of-work. Bitcoin (being forked now) doesn't and ethereum (already forked twice) definitely doesn't.

Blockchains are a social construct and what ultimately protects them are the participants. Exactly like actual currencies they are subject to market forces; people either believe in and want the currency or they don't. It's this confidence that gives the currency strength.

More simply:

"On medium to long time scales, humans are quite good at consensus. Even if an adversary had access to unlimited hashing power, and came out with a 51% attack of any major blockchain that reverted even the last month of history, convincing the community that this chain is legitimate is much harder than just outrunning the main chain’s hashpower. They would need to subvert block explorers, every trusted member in the community, the New York Times, archive.org, and many other sources on the internet; all in all, convincing the world that the new attack chain is the one that came first in the information technology-dense 21st century is about as hard as convincing the world that the US moon landings never happened. These social considerations are what ultimately protect any blockchain in the long term, regardless of whether or not the blockchain’s community admits it (note that Bitcoin Core does admit this primacy of the social layer)." [1]

[1] https://medium.com/@VitalikButerin/a-proof-of-stake-design-p...

The subtlety here is making the distinction between confidence and resistance to attack. They're not the same thing at all. And what's interesting here is that, going off the model of real currencies, one could assume market forces will prevail here. Developed nations let their currencies float precisely because on the whole they're confident that the market will ultimately punish any malicious attacker who attempts to destroy the currency by selling large amounts of it.

Re: Why aren’t distributed systems engineers working on blockchain technology?

#115

Post by ty_a »

matt_wulfeck wrote: post #41
aphextron wrote: post #3

>The blockchain community is immature and historically unwelcoming. >It feels like a get rich quick scheme - https://twitter.com/naval/status/878018839044161536 This is what it comes down to IMO. Anything cryptocurrency related is just surrounded by a cloud of shady characters and scammers, even though the tech is legit. You just can't trust anyone, nor their intentions. Living your life and working in a constant sta…

Even if the tech is "legit" it's encredibly inefficient. We're spewing off tons of CO2 just to verify the block chain, but we have plenty of technologies that can do the same with a minuscule amount of energy. The problem is that they want a system that's trustless and that's simply not what most engineers care about. We want a system that's fast and efficient.

Most mining is hydro powered.

Re: Why aren’t distributed systems engineers working on blockchain technology?

#116

Post by KirinDave »

ty_a wrote: post #115
matt_wulfeck wrote: post #41

Earlier quoted context omitted.

Even if the tech is "legit" it's encredibly inefficient. We're spewing off tons of CO2 just to verify the block chain, but we have plenty of technologies that can do the same with a minuscule amount of energy. The problem is that they want a system that's trustless and that's simply not what most engineers care about. We want a system that's fast and efficient.

Most mining is hydro powered.

Please cite your sources.

Beyond this: https://www.coindesk.com/carbon-footprint-bitcoin/

Which is sort of unvalidated and points to one or two actors in the space suggesting that they want to move to non-fossil fuels and one that miiiiight be.

Re: Why aren’t distributed systems engineers working on blockchain technology?

#117

Post by svara »

vocatus_gate wrote: post #105
srdev wrote: post #99

My observation is this: The "killer app" of Blockchains are the trustless nature of the system. In most practical application, allowing actors without a trust relationship to interact with the log is not a significant requirement of the system. Once you remove that requirement, there are technically superior approaches to maintaining a ledger.

That's why I've never understood all this talk about "private blockchains." Why? What's the point? We already have distributed database systems that are light years faster than blockchains. The ONLY benefits of a blockchain-based approach are censorship resistance and not having to trust the actors in the system. If you don't need either of those two benefits, then there are far better existing solutions for record k…

As far as I've seen, the point is mainly PR. These private blockchains are usually not bitcoin-style proof of work chains, but would more accurately just be called distributed databases. For example, look at https://www.openchain.org/, "How does it work" section. A lot of altcoins actually are the same, with explicit "trusted nodes", such as Stellar (https://www.stellarcoin.org).

Happy to learn about any counterexamples and use-cases for those.

Re: Why aren’t distributed systems engineers working on blockchain technology?

#118

Post by SCHiM »

DINKDINK wrote: post #98
roywiggins wrote: post #38

Earlier quoted context omitted.

And if there's a fork? As I see it, either you follow the fork supported by the majority of miners, or you run the risk of finding your coins worthless, because everyone else is on the other fork.

>And if there's a fork? If someone proposes to violate the rules of your blockchain, you ignore them. Only when users support rule changes does the market value them. >either you follow the fork supported by the majority [...], or you run the risk of finding your coins worthless Any thing is only worth what others will pay for it. This is the case regardless if you are using a blockchain or not. If the world suddenly…

This is true in the abstract. However in the real world Bitcoin is in such a bind right now. While the USD had been stable in that sense since it came into existence.

Re: Why aren’t distributed systems engineers working on blockchain technology?

#119

Post by icebraining »

sigstoat wrote: post #46
rafiki6 wrote: post #39

Here's one take DLT and Blockchain: https://hbr.org/2017/01/the-truth-about-blockchain TL;DR, it's gonna take a real long time to make blockchain useful, if it even is. Further, I have yet to see a real production application of blockchain that isn't a crypto-currency. Everyone and their grandmothers has invested in it, or started a company or w/e. But has it actually been used to solve another problem? Blockchain wa…

> But has it actually been used to solve another problem? Maersk is interested http://fortune.com/2017/03/05/maersk-tests-blockchain-based-... Makes some sense freight shipments involve a lot of parties who don't completely trust one another, and they've got a lot of information to move back and forth between one another.

The innovation of Bitcoin is combining the Proof-of-Work with a signed ledger. But Proof-of-Work is only needed because otherwise anonymous people might create new nodes from thin air and get a majority.

In this case, all the parties are well identified, so if a new node appears, it can simply be rejected by the existing participants. So there's no need for PoW, all you need is for companies to sign and publish documents to each other. The only possible fraud is publishing different documents to different parties, but that's easily fixable by having the nodes confirm each others' documents.

So yeah, that's just a signed ledger. All invoicing programs in my country had that before Bitcoin even appeared, as part of the SAF-T standard (mandated by our tax authority). It's nothing new.

Re: Why aren’t distributed systems engineers working on blockchain technology?

#120

Post by tekromancr »

indigochill wrote: post #94
billmalarky wrote: post #64

Earlier quoted context omitted.

The Ethereum team is serious about proof of stake.

Ethereum has not shown a particularly strong track record when it comes to security, so I think the comment stands until they address that or get replaced by the next hotness.

I think that statement is a little disingenuous. There haven't been any attacks that target Etherium proper, only bugs in smart contracts riding on top of Eth.
Post reply on HN