Why aren’t distributed systems engineers working on blockchain technology? • Hacker News Board

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

#171

Post by gwbas1c » Tue, Aug 01, 2017, 8:59 PM UTC

JohnJamesRambo wrote: post #140

gwbas1c wrote: post #138

I work in distributed systems, and I consider blockchain very interesting. It's so interesting I click on a lot of links on Hacker News, and sometimes I take an afternoon on a weekend or vacation to read through some of the papers. I just have a well-paying job and I don't see any opportunities lucrative enough for me to jump away. Why? Bitcoin is an experiment. Reading through the paper, Bitcoin does not scale to ge…

What are the real uses for blockchain you could see?

I see it going in two different directions:

Blockchain: Useful where there are groups of actors, (people, countries, companies, ect,) of a large but limited size, incomplete trust, and limited exchanges. (For example, a group of 1000) A blockchain could be used to publish who owns an asset; as long as the blockchain technology can scale to the number of assets, number of owners, and number of transactions that occur. The actors need to have sufficient motivation to provide the computing power to run the blockchain; otherwise, it makes a lot more sense to just pay a private clearinghouse.

The problem with "blockchain" is that every computer in the network has to keep a complete copy of the ledger. Useful in the above examples, but this is also highly limiting to scalability, which leads to...

Blockchain-like: I think "blockchain" will lead to blockchain-like technologies where every computer doesn't have to keep an entire copy of the ledger. Instead, the graph of computers allows for searching publicly published knowledge, and participants are encouraged to preserve facts that are in their best interest. This allows for much higher scalability because computers in the network can merely ignore most transactions. If you are a small player in the network, you might pay a larger player to monitor the network and provide you with search.

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

#172

Post by davidgerard » Tue, Aug 01, 2017, 9:01 PM UTC

darawk wrote: post #131

jethro_tell wrote: post #128

Earlier quoted context omitted.

No that's not illogical, Why would I work to bend something like block chain, that was designed to be inefficent, to my will when I can just start from scratch using projects that were designed to be efficient first? That's very logical, the kind of decision that an engineer that's getting paid the make the best distributed system for the least money would make.

It's not designed to be inefficient. It's designed to solve the distributed consensus problem in an adversarial environment. There are other now proposed mechanisms of doing that (e.g. PoS) that are quite efficient. Working on those would be extremely beneficial to the cause of environmental efficiency.

> There are other now proposed mechanisms of doing that (e.g. PoS) that are quite efficient.

Well, no they aren't, because they're proposals.

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

#173

Post by gwbas1c » Tue, Aug 01, 2017, 9:01 PM UTC

ProblemFactory wrote: post #152

gwbas1c wrote: post #138

I work in distributed systems, and I consider blockchain very interesting. It's so interesting I click on a lot of links on Hacker News, and sometimes I take an afternoon on a weekend or vacation to read through some of the papers. I just have a well-paying job and I don't see any opportunities lucrative enough for me to jump away. Why? Bitcoin is an experiment. Reading through the paper, Bitcoin does not scale to ge…

> Someone needs to pay me to implement something that requires blockchain It's difficult to find something that requires a blockchain, because the unique distinguishing feature of blockchains is the lack of any trusted central authority. If you have a trusted authority, then you can do anonymous currency, money transfers, smart contracts, or anything similar 1000x more efficiently with a single server, an API, and so…

Which makes it look like blockchain is a solution looking for a problem.

I still believe that problem will come up, and/or we'll see technologies that borrow concepts from blockchain without being true "blockchain."

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

#174

Post by davidgerard » Tue, Aug 01, 2017, 9:05 PM UTC

tekromancr wrote: post #120

indigochill wrote: post #94

Earlier quoted context omitted.

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.

> 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.

This is a bit No True Scotsman. It's like "secured by math!" about Bitcoin - the ecosystem is rife with fraud and error.

So too with Ethereum. Given that smart contract functionality is literally Ethereum's unique selling point, you can't claim that is somehow not something that can be discussed as a problem with Ethereum. They broke the immutability guarantee for one bad smart contract, after all.

You really can't say "Ethereum is completely secure!! Except for everything that people actually use it for" without being more than a little disingenuous yourself.

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

#175

Post by davidgerard » Tue, Aug 01, 2017, 9:09 PM UTC

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.

Actually, Bitmain's building with 75% of mining capacity appears to run on wind and solar.

https://bitcoinmagazine.com/articles/bitmain-reveals-plans-f...

(was planned then, I believe it's in service now)

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

#176

Post by darawk » Tue, Aug 01, 2017, 9:09 PM UTC

davidgerard wrote: post #172

darawk wrote: post #131

Earlier quoted context omitted.

It's not designed to be inefficient. It's designed to solve the distributed consensus problem in an adversarial environment. There are other now proposed mechanisms of doing that (e.g. PoS) that are quite efficient. Working on those would be extremely beneficial to the cause of environmental efficiency.

> There are other now proposed mechanisms of doing that (e.g. PoS) that are quite efficient. Well, no they aren't, because they're proposals.

No they're not. They're implemented in many altcoins. And to say that 'proposals' are not efficient because they haven't been implemented is both semantically wrong, and an incredibly stupid point even if it weren't.

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

#177

Post by darawk » Tue, Aug 01, 2017, 9:12 PM UTC

dogecoinbase wrote: post #169

darawk wrote: post #165

Earlier quoted context omitted.

> Name a problem that would best be solved with blockchain. Money. E.g. Bitcoin.

Bitcoin as money has all the worst properties of both credit cards (traceable, non-anonymous) and cash (irreversible, theft-prone).

It isn't intrinsically theft prone, the UI around using it is just not up to snuff yet. Credit cards, on the other hand, are intrinsically theft prone. They are a pull architecture. Every time you use your credit card, the store takes money from you. This is patently absurd, and the only reasonable definition of 'theft prone'. Bitcoin is a push architecture. You can never take my bitcoin, I can only send you bitcoin. Further, irreversibility is definitely what you want in the underlying protocol. It's fine if people want to implement escrowing and chargebacks via intermediaries on top of Bitcoin, and I would encourage that. But at a fundamental level, a payment system should be irreversible, immutable, uncensorable and push-based. Which is exactly what Bitcoin is.

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

#178

Post by jliptzin » Tue, Aug 01, 2017, 9:36 PM UTC

jethro_tell wrote: post #136

jliptzin wrote: post #122

Earlier quoted context omitted.

This is true of any new burgeoning technology. Were you around for the Internet boom in the 90s?

I keep hearing this is like the internet from 30 years ago. Thirty years ago, there was no good way to do ecommerce or send an email. Today, there is a good way to pay for things and do all financial transactions in a simple low hastle way. The internet was sending instant, asyncronious communication vs letters (async) or phone (instant). Bitcoin is creating phone calls and post mail in the email age and using a meta…

I was talking about how scam artists flocked to the internet with pyramid schemes, chain emails, get rich quick schemes, etc. All that stuff is still here, but it's drowned out by people doing actual normal business now.

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

#179

Post by darawk » Tue, Aug 01, 2017, 10:59 PM UTC

davidgerard wrote: post #174

tekromancr wrote: post #120

Earlier quoted context omitted.

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.

> 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. This is a bit No True Scotsman. It's like "secured by math!" about Bitcoin - the ecosystem is rife with fraud and error. So too with Ethereum. Given that smart contract functionality is literally Ethereum's unique selling point, you can't claim that is someh…

It is, thus far, completely secure. The fact that SQL injection bugs exist does not negate the security of HTTP. The fact that stack overflow bugs exist does not make x86 insecure. The primitives you are provided are secure - how you use them may not be.

Now you may legitimately argue that Ethereum is poorly designed to encourage secure contract authorship. And I would agree with you in many respects. But that is a distinct concern from the security of Ethereum itself. Conflating the two is at best confusing and at worst maliciously spurious.

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

#180

Post by ExactoKnight » Tue, Aug 01, 2017, 11:49 PM UTC

jethro_tell wrote: post #150

darawk wrote: post #131

Earlier quoted context omitted.

It's not designed to be inefficient. It's designed to solve the distributed consensus problem in an adversarial environment. There are other now proposed mechanisms of doing that (e.g. PoS) that are quite efficient. Working on those would be extremely beneficial to the cause of environmental efficiency.

Evironmental efficiently solving a problem I don't have is still inefficient. I still can't think of a mojor user for proof of work in a destributed system. I have trouble thinking of a distributed system that someone would pay for to not be centrally trusted, usually by the benefactor of employment? We always hear what a great solution blockchain is. Even if that's true, I never hear the actual problem it solves. Su…

The block chain offers the following unique benefits for asset transfer:

- Assets move directly between entities — transfer and settlement are one-and-the-same — eliminating counterparty risk (https://en.wikipedia.org/wiki/Credit_risk#Counterparty_risk)

- Assets move instantly, and settlement time is ~10 minutes (settlement can also be instant if the parties trust the other won’t “double spend”)

- The transfer mechanism is public infrastructure and is always improving — everyone has a shared incentive to make it better, similar to Internet protocols like email and http

- Assets and end-user data are privately controlled, and strong security for assets can be achieved by using multiple signing keys across several parties

- Policy rules about the movement of assets can be enforced programmatically — whether those are “terms and conditions” or regulatory requirements

- Assets are fungible and play nice together — e.g., you can use reward points to buy mobile minutes

- A single transaction can include multiple entities and assets, on both sides of the transaction — e.g., you could execute a merger of two companies in a single transaction, with the inputs to the transaction being all stockholders across all share classes for both companies, and the outputs being all the newco shares going to all the new stockholders (again, no escrow service needed)

- Every transaction is added to an immutable record which, while anonymous, can be used to construct a perfect audit trail of an asset’s movement when combined with the private data held by the entities using the system — this defends against fraud, and also gives issuers transparency into asset movements

- It’s easy to integrate new parties

- Like the Internet itself, it’s a global system

Worth a read: https://blog.chain.com/wall-street-meet-block-chain-b2747909...