Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

It's striking how much the crypto world depends on trust in other parties. The whole point of crypto was supposed to be that it was "trustless". But it's not set up that way. All these crypto derivatives are not set up as contracts on a blockchain, with assets locked up until the derivatives settle. They're book entries with some weakly regulated exchange in Outer Nowhere.




The people who want trustless decentralization and the people who want leveraged gambling and the people who want KYC-free international money transfer may be different people. The only problem with Liberty Reserve was that it got shut down; if a "decentralied" fig leaf can allow it to operate... let there be "decentralization".

> The whole point of crypto

There is a common confusion in this (perhaps?). Most businesses get created primarily to make money. Not primarily to solve the world's problems. It's easy to say "if they really had their customers at heart...". Well, yeah, but that's not and has never been the priority. It's not a cynical view, it's being realistic.

All kinds of mayhem follows. All the way to fundamental research papers such as "on average actively managed mutual funds do not beat XX index". Well, yeah, mutual funds don't get created because someone is good at it. They get created because someone wants to make money. Beating XX is not the first objective, or competence, of the entrepreneurs. Hopefully that fund doesn't last too long but often it does, and anyway there are many of them.

So anyway, there are plenty of ways to try and leverage ideas of cryptography, crytocurrencies, block chain - most of which are still accessible - and most of the ventures in the field are not going to be primarily about solving the users' problems.


You can trade perpetual futures, onchain, mostly decentralised, in self-custodial manner [1] e.g. on GMX

https://gmx.io/

Some more modern decentralised exchanges (DEXes) dealing with leveraged trades and try to minimise centralisation also include YieldBases:

https://yieldbasis.com/markets

There are other exchanges that are much more centralised, like Hyperliquid, and it is incorrect to call these decentralised. But there are truly decentralised alternatives as well.

GMX is not as popular, let's say Binance, because onchain user experience has been very hard. You don't want to sign every order from your crypto wallet. Transaction cost ("gas fee") used to be too high for trading. This is finally changing with the latest Ethereum improvement proposals, dealing with so called account abstraction.

[1] Because futures always settle on an external price, the price feed must come from some oracle. In the case of GMX, there are keepers (multiple of them) who are responsible to bring the correct price to Arbitrum chain and trigger the settlement. But it's not a single party.


That's not true with decentralised exchanges like hyperliquid, no?

Hyperliquid and similar exchanges aren't decentralized. That is their long term goal but they are very far from achieving it.

The few actual decentralized exchanges are too slow and expensive.


There are some exchanges that are more decentralised (and older) than Hyperliquid. Hyperliquid, while being the most popular one, is not the only horse in the town.

E.g. GMX on Arbitrum chain is no longer prohibitively expensive.

Left some comments here https://news.ycombinator.com/item?id=46172450


I mean, as soon as synchronisation is required in any system, block chain, distributed SAAS, even Peer to Peer sharing, decentralisation fails hard

That's one of the sticking points I have with the /idea/ of the technology


Ethereum and similar chains run arbitrary computation on-chain. You can make a futures exchange on Ethereum (or Solana, etc). However, the fees for doing so are very large, and confirmation times are very long, like any other on-chain transaction.

What I am loving about this comment, and the downvotes, is the idea that blockchains can escape things like basic laws of the universe.

> confirmation times are very long, like any other on-chain transaction

Yes. synchronisation is where everything breaks down because you have to get everyone to agree to the new state.

edit: Sorry, not everyone, but a consensus, and that consensus is then what everyone agrees is the state.


> HyperCore includes fully onchain perpetual futures and spot order books. Every order, cancel, trade, and liquidation happens transparently with one-block finality inherited from HyperBFT. HyperCore currently supports 200k orders / second, with throughput constantly improving as the node software is further optimized.

Key part:

> fully onchain perpetual futures and spot order books


Being on a blockchain and being decentralized are two different things. The HyperCore client isn't even open source.

That's just patently false.

> Importantly, HyperCore does not rely on the crutch of off-chain order books. A core design principle is full decentralization with one consistent order of transactions achieved through HyperBFT consensus.


The basis of decentralized software is open-source. Otherwise a centralized authority can just push an update to, for instance, blacklist addresses.

https://github.com/hyperliquid-dex/node

"For lowest latency, run the node in Tokyo, Japan."

Decentralization means to run all of the closed-source nodes in the same AWS datacenter!


And in fact they did just this when their vaults started bleeding money on an unfavourable position (JellyJelly). They handed out a closed source binary and the validators ran it immediately, closing out the market at an arbitrary price.

The basis of decentralized software is open protocol. Then it doesn't matter that somebody runs closed source while somebody runs open source.

as an operator you don't even get the real validator / node binary directly, nor can you control which version to run.

all you can do is run their visor, and they push out whatever proprietary blob they produce and restart "your" nodes at their command.


Different cryptocurrency products offer different properties and guarantees. Much like different databases offer different concurrency models. Folks that use currency backed stablecoins do not care for the trustless properties. There are various algorithmic stablecoins out there that you can use to stay free of KYC/AML but they aren't very popular.

Largely the folks that want trustless currency use chains like BTC, BCH, XMR, or ZEC.


This comment makes sweeping generalizations.

This is a common place in any thread about cryptocurrencies on HN unfortunately... I could be convinced of my own message also being a sweeping generalization if anyone can point out a single post where top comments aren't doing exactly this when it comes to this topic, even the technical ones.

This is a useless comment for most readers (myself included!) unless you specify what you think those sweeping generalizations are, and why you think they're unsound.

It was the case up until recently. But today Hyperliquid does it on chain and very popular.

Hyperliquid being on chain in the traditional sense is fiction. You have a closed source piece of software run by closely controlled "validators" with additionally centralised components.

By now crypto-in-practice has violated so many of its supposed founding principles that it's tired and cliche to point it out.

It was supposed to be limited in supply unlike fiat, and yet Tether underpins the whole thing and they print that out of thin air all the time. It was supposed to be decentralized, but in practice a few big exchanges control all the transactions and a few big mining pools control all the minting. It was supposed to be "code is law", and yet if you find a big exploit on smart contracts it'll be unwound later on and the cops will still show up for you. And as you say, it was supposed to be trustless, but counterparty risk is everywhere.

And it turns out nobody cares, because to a first approximation nobody is in crypto for the libertarian principles. It is all about number go up; always has been, always will be. It's not even worth pointing out anymore.


> It was supposed to be limited in supply unlike fiat, and yet Tether underpins the whole thing and they print that out of thin air all the time.

This is a joke right? Tether (USDT) is pegged to the dollar... and there is not really a limit to the USD printing machine, nobody ever claimed a stablecoin would have a limited supply. It's literally the main critique of the fiat system levied by crypto proponents.

The only asset which has made and still hold promises of not increasing its supply over its limit set through its consensus code is Bitcoin. And it is nowhere close to ever change... as a matter of fact if it changed, most people wouldn't call that fork Bitcoin.


The problem with Tether is that they are tight-lipped about their backing assets. No one knows if the peg is real, it's just "trust me bro"

Well they publish attestations from third-parties, but no full audits, so sure they could be much more transparent.

But the claim about USDT ever claiming that its supply wouldn't increase is pure fantasy. It literally makes no sense if you understand how the peg is maintained (technically by minting and burning tokens).


I took their comment to mean that tokens valuations are tied to stablecoins. Sufficiently tied enough as to be de facto properties of tokens themselves.

Also interesting that tether is the private largest holder of gold at 14 billion $ xaust

Yes. Cryptocurrencies operate within the larger econo-political system we live in, and as long as cryptocurrencies replace only a part of that system, the rest of the system will continue to operate as it does otherwise. Its quite clear that the way capitalism operates in practice is that most markets end up being oligopolish, and that people with guns are needed to keep the system stable. So not at all surprising.

> And it turns out nobody cares, because to a first approximation nobody is in crypto for the libertarian principles. It is all about number go up; always has been, always will be. It's not even worth pointing out anymore.

I agree 100% - Meme stocks go brrrrrrrr

The idea that it's a currency that lives beyond the reach of governments is laughable (as soon as something goes bang a lot of the owners call for... regulators and government oversight)


> I agree 100% - Meme stocks go brrrrrrrr

Mostly, meme coins go into a screaming dive after the initial pump. Go type some meme coin names into Coinmarketcap.

Except for Bitcoin and Ethereum, almost everything in crypto has crashed hard.


People putting their self-interests before maintaining support for more general principles is par for the course.

Even the vast majority of free-market maximalists will support a government bailout of large banks or the auto industry if it will save their investment portfolio.


whats more important to me is that you don't have to ask anybody if you can deploy an entire financial services suite

and not only will other people worldwide use it immediately, they will also pay for all your infrastructure costs as they update the chain state with every transaction fee that they pay

the permissionless nature means you can deploy anything as cenralized or decentralized as you want, and its up to consumers to be discerning and its only their fault if they are not

cost wise this will always be attractive to developers and for them to bring over every audience they can muster, because web 2.0 cloud cannot compete with that cost structure and permissionless nature


Isn't basically virtually 100% of the money that isn't crime adjacent web 2.0 implying it can compete?

What are you asking? Can you rephrase that in a different way?

Can I leverage trade derivatives and also earn fees from liquidity pooling with Robux?


BTC almost exclusively enables crime. It's fundamentally too bad at basically everything to replace any part of the real economy. It is almost exclusively used for crime, admittedly fun technological exploration, and gambling on a valuation based not on actual net utility in current context but on perception of future utility that will probably never materialize.

Web 2.0 based on boring old primitives like ad dollars and banks actually funds things that in real life provide ultimately virtually all the actual utility obtained by the world from software.

You said

> web 2.0 cloud cannot compete with that cost structure and permissionless nature

It appears to me that that is just incorrect on its face because web 2.0 cloud actually DOES compete insofar as its literally everywhere as we speak and web 3.0 is a buzzword from 2014 that has yet to achieve actual meaning.

To rephrase do you feel it is accurate to say that something that represents basically all the real value obtained by network computers doesn't competes with something that provides? What again?


BTC isn't a programmable chain. As yieldcrv says, your talking points are too old to make sense anymore. BTC's big "innovation" has been its version of an L2: the Lightning Network. Otherwise BTC people mostly treat it as a gold-like asset at this point.

BTC doesn’t really operate in the decentralized finance world, all the platforms for applications are on other chains so your arguments are also stuck in 2014

Ask an AI about it to catch up, this is a decade too late to have that conversation

the only thing that matters is that there is liquidity and permissionless deployment, we are far far beyond “should there be liquidity”, you can build business on smart contract platforms solely because there is liquidity and people with frictions you can solve just like any other industry or the financial services sector in general




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: