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

Why blockchain is not yet working (2018)

as1ndu.xyz

Post reply on HN
Print view View original

301–310 of 330 posts

Re: Why blockchain is not yet working (2018)

#301

Post by jobigoud »

xirbeosbwo1234 wrote: post #272
zeroxfe wrote: post #228

Earlier quoted context omitted.

A Merkle tree is just one component of a blockchain. It also requires: a byzantine consensus mechanism, a p2p network stack, public key crypto, and a set of economic incentives to keep it going. You can't build a public database on a Merckle tree without being able to agree on what goes into it.

That is a cryptocurrency. The work coming from the people at Hyperledger, who are also using blockchains, is an ordinary database complete with access controls. It uses the same data structure but without any of the political goals and with many different design decisions. A blockchain is just a data structure. It isn't even a very interesting one. In common parlance "blockchain" isn't well-defined: it just refers to…

You can't remove the consensus mechanism part and still call it a blockchain.

Re: Why blockchain is not yet working (2018)

#302

Post by rini17 »

AngrySkillzz wrote: post #215
rini17 wrote: post #47

As the recent GameStop events have shown, it's uncertain who owns how many shares, settlement is slow and error/bezzle-prone, etc. etc. Plus I venture to guess that HFT algorithms do consume country-level amounts of energy as well. Unlike blockchain, it's not surplus electricity from renewables in faraway places. And all that energy does not even improve the whole ledger security. But hey, unlike the filthy blockchai…

Go read Money Stuff before you spread more misinformation about how the financial system works. There are no counterfeit shares, ownership is obvious except for cases like voting (because a short can pay a negative dividend but can't perform a negative vote), settlement is not error-prone and is slow on purpose (Russia tried real-time settlement and went back to T+2), HFT is literally just a service provider doing in…

I am confident we'll know more in few years. Something being merely illegal has never stopped anyone.

Re: Why blockchain is not yet working (2018)

#303

Post by ENGNR »

nexthash wrote: post #11

I think the #1 reason blockchain doesn't work is very simple - the technology is broken beyond fixability. This article is two years old, but the same variations of these excuses always come up when the crypto ideologues are grilled on why their utopiaic vision hasn't come about, for all of blockchain's existence. Blockchain is awful for most things compared to a centralized or distributed database: it is inefficient…

I think the real innovation is permissionless

Yes you could get permission.. but it would take time. What if you want to do some random thing over a weekend? You’re not going to want to ask permission

Asking permission also opens the giver up to an implicit risk. Not just in a legal sense, but if you need to ask a mid level manager for permission they have nothing to gain and everything to lose

Putting everything out there on a permissionless chain with a rock solid set of predefined rules, it’s the hacker dream

Re: Why blockchain is not yet working (2018)

#304

Post by mekster »

nexthash wrote: post #11

I think the #1 reason blockchain doesn't work is very simple - the technology is broken beyond fixability. This article is two years old, but the same variations of these excuses always come up when the crypto ideologues are grilled on why their utopiaic vision hasn't come about, for all of blockchain's existence. Blockchain is awful for most things compared to a centralized or distributed database: it is inefficient…

I'm almost about to think people who just keep criticizing blockchain on wrong points are those who missed the investment opportunity to just get the steam off.

Blockchain on Bitcoin is meant to be inefficient. That's why there's a fair distributed race because it's a waste and no one has any head start or another motivation to mine to go ahead of the others.

You can come to think of something efficient and you should be able to start your blockchain and get funded well but criticizing is certainly the easy part.

Re: Why blockchain is not yet working (2018)

#305

Post by TheColorYellow »

DennisP wrote: post #87

The article mentions Ethereum several times, so here are some Ethereum-specific updates for 2021. It's still a work in progress but some progress has been made. 1) Scalability For now, the big advance is a layer-2 approach called "rollups," several of which are in production already and able to handle several thousand transactions per second. Zkrollups do that without compromising security assumptions at all, but are…

This is an excellent response and I wish it was a bit higher up for people to see.

My only retort to your points would be that much of what you've described is still in nascent stages or at best, early go-live. I'm rooting hard for Ethereum to get ahead of some of these problems, but with all the noise in the community the last few years, it's difficult to determine what is actually legitimate and what is simply being oversold.

My hope is that the Defi boom will continue it drive attention and resources to the space that will help to realize some of this potential. Despite that, at the moment I still can't help but be a skeptic myself.

Re: Why blockchain is not yet working (2018)

#306

Post by mycall »

GuB-42 wrote: post #257
hannob wrote: post #114

Earlier quoted context omitted.

So Blockchain is a fancy word for something like a Git repository?

That's a question I asked myself a few times. The fundamental part of a blockchain is that each block signs the previous block, making it very difficult to rewrite history. And it is exactly what git does. We could implement a cryptocurrency on a git repository. Just commit and push 'I am X, here is the proof, send Y gitcoins to Z' whenever you want to do that. The only thing git is lacking compared to say, bitcoin,…

Perhaps GitCoin should be a thing.

Re: Why blockchain is not yet working (2018)

#307

Post by mycall »

jnwatson wrote: post #137
davidgerard wrote: post #94

Earlier quoted context omitted.

> But for most newer projects "blockchain" just means an auditable trail of state changes. you mean, the 1979-model Merkle tree? I concur that in a lot of marketed "blockchains", the closest to a useful bit is the 1979-model Merkle tree.

Even more trivial, single-leaf Merkle trees is all that is needed.

Even less if you look at NANO and IOTA.

Re: Why blockchain is not yet working (2018)

#308

Post by lapinot »

StavrosK wrote: post #227
zaphar wrote: post #221

Earlier quoted context omitted.

Neither of your examples are tradeoffs. Both my computer and my phone are excellent at their jobs with little in the way of downsides. Bitcoin is bad at it's job with lots of downsides. It only looks better than the alternatives if you live in a place where the economy is heavily mismanaged. If you're currency is only better than the absolute worst economies in the world don't expect all the people in better economie…

You keep moving the goalposts between "blockchain the technology" and "Bitcoin the currency". If you want a trustless auditable log, your best option is a blockchain. What's a better alternative?

There is nothing "trustless". Our world is social by structure, all our actions contribute to various kinds of trust. Blockchains aren't trustless they're just unable to internally express trust relationship: there is always trust in the devs, in the exchanges, in the miners, in the other holders. There are political events like "blockchain forks", protocol upgrade, parameter changes..

So much for the trustless auditable log. Btw "auditable log" is easy (eg certificate transparency), the hard part is deciding what goes in it.

Having said that you always trust something, you better do it consciously and run some real consensus algorithm like raft or stellar consensus protocol.

Re: Why blockchain is not yet working (2018)

#309

Post by pavlov »

centimeter wrote: post #296
pavlov wrote: post #287

Earlier quoted context omitted.

I'm genuinely curious where you're sending money that a wire transfer using IBAN+SWIFT doesn't work but Bitcoin does. What country has the banking infrastructure to support easy $50k-equivalent withdrawals from a local Bitcoin exchange, but not for receiving wires?

If you don't want to get screwed on FX, you need to use a brokerage intermediary with entities in both currency domains. Direct international wire usually causes problems. So if I'm sending FX, it looks like US Bank -> Wire -> IBKR -> Wire -> SG Bank (or whatever) Usually this takes at least 5-9 business days end-to-end. (Wire delay, withdrawal hold, wire delay on the other end.) A bitcoin transaction takes a few min…

It doesn’t make sense that you don’t want to pay FX rates but would happily accept the payment in Bitcoin whose rates fluctuate much worse.

Some banks charge up to 2% for currency conversion, but surely that’s better than the 10-15% lottery involved in Bitcoin? Unless you want to have a gambling aspect to your invoices.

Re: Why blockchain is not yet working (2018)

#310

Post by pavlov »

selestify wrote: post #298
pavlov wrote: post #295

Earlier quoted context omitted.

What’s the benefit? A stablecoin is like an account at a weird unregulated bank from where you have to move the money to a real bank to use it. How does using one of these products let me access the money faster than a wire between the sender’s USD account and mine?

Wire transfers take multiple business days, at least in the US, at least for me every single time I do one. Individual banks may also put restrictions on your domestic wire transfers. Maybe it's different for you since you're a business, but as an individual, there's a lot of friction in sending money to my friends. > unregulated Depending on your perspective, this is either a feature or a bug. In my opinion, a big f…

To be clear, this is primarily a US problem.

I live in New York now, but previously I lived in Europe and UK. Bank transfers there are instant. If you want to send 20k EUR from Finland to Spain, it can be done immediately, 24/7, with minimal fees. Same in UK for GBP transactions.

So it’s perfectly possible to do this without a cryptocurrency. It takes some political will though. But then again, it also takes political will to prevent cryptocurrency exchanges from being shut down if they truly threatened central banks.

US wire transfers seem to depend entirely on goodwill of banks. I regularly send wires between Wells Fargo and Schwab, and they always complete in a few hours. Yes, it sucks that banks get away with bad service. But routing money through unregulated cryptobanks isn’t a long-term fix because the regulators will come for them.

Post reply on HN