Hacker News new | ask | show | jobs
by Dwolb 1618 days ago
Curve Finance with $14B locked to its platform. Serves up low slippage stablecoin swaps.

Some background on Curve (and Convex): https://medium.com/coinmonks/convex-curve-curve-d7e28cd6c1d9

2 comments

“Low slippage stable coin swap” is one of the most jargony things I’ve read on here in a while. Can you please ELI5 because I have no clue what this is or how it could be useful.
Have you heard of leverages, margin calls, derivatives, put options, etc.? DeFi is pretty much the same level of complexity but implemented on top of cryptocurrencies.
This https://youtu.be/8CAafjodkyE?t=104 is a recent thing I found mind blowing; 2020 iPhone can do on-device image processing and object recognition and speak a description of what the camera sees. Just quietly a builtin feature.

2010 iPhone, I was mind-blown by the Word Lens app which could do OCR on text in the camera feed, translate the words into another language, and overlay the results on the image in near-realtime. http://edition.cnn.com/2010/TECH/mobile/12/20/word.lens.ipho...

These are tasks I had never seen any computer do in any circumstances, things which would be unthinkable on a Java MIDP Blackberry from 2005, or a Pentium II desktop from 1997. Compare this to the wearable augmented reality devices Professor Steve Mann was building through the 80s and 90s[1] and this "describing the image" is so so far ahead in so many ways - processing power, imaging quality, battery life, storage space, size, weight, convenience, reliability, it's just mindblowingly better, neural networks and fast chips and solid state storage is a step change in a way that "faster" isn't enough to really convey.

Google tells me "Derivatives have a long history in the United States, dating back to the founding of the Chicago Board of Trade in 1848.", you're telling me people are doing that but with blockchain, why is that interesting at all, why mind blowing?

[1] http://cyborganthropology.com/Steve_Mann

Personally the reason it blows my mind is because it’s now permissionless to craft a unique structured product or derivatives protocol, deploy it globally, and anyone on the planet can participate in that market.

It’s fully customizable and accessible to anyone with an internet connection.

Call me old fashioned my I prefer regulations in the markets I participate in. I still fail to see how “permissionless market creation”, while impressive, actually solves a problem. Where is the demand coming from, besides people in tech??
In what way "permissionless"? In the way "the government has agreed that they can be unregulated financial products" or in the way "hopefully it can evade government financial regulation"? or in some other way?
Sure there’s definitely US-centric issues given the current regulations in place.

But it’s permissionless in that if the financial markets don’t support a structured product you or your community needs, you can create and deploy it yourselves with instant global accessibility.

Evasion of regulation. Not just from governments, however. From anybody who wants to control people without consent.
That’s not a great argument. It’s like comparing the invention of tcp with the iphone. They’re just two different technologies. If you’re interested in learning there are many resources, but it’s too easy to criticize something by looking at its cover.
And if you told me there was a new "TDP/JP" which is like packet switching but much slower and more energy hungry, and it was blowing your mind, and I asked why, and all you could say was "do your own research", I don't think I would do any research about that either.

Money obsessed people moving money around in ways that seem manipulative and gambling feels like a negative influence on the world.

Ok cool, so basically just fancy ways for traders/investors to gamble and “create value” that doesn’t really benefit anyone besides themselves and others tied to the platforms. Got it.
The idea is you can lock capital in a pool, and as people trade between two usd stablecoins, USDC and DAI, you earn a fraction of the trade fees. The low slippage is due to the pool being large enough that you will get $1->$1.
So this is Yet Another Liquidity Pool? $14B is not the kind of scale I mean. That is frankly trivial especially when transactions are large (or initial funding is high).

What I mean is, if blockchain is going to be a revolutionary technology backing all finance, where is the evidence it can handle the kind of transaction volume that, say, major credit card networks generate?

Maybe check out DYDX L2.
Great pointer, thanks.

This is a protocol that rolls up ETH transactions and executes them in batch, in an attempt to alleviate scalability concerns. I can't really speak to the side/unintended effects of transaction rollups on a blockchain, but I'm interested in learning.

This excerpt from the FAQ

> It is worth noting that anyone can become a relayer so long as they have staked the required bond in the smart contract. This incentivises the relayer not to tamper with or withhold a rollup.

kind of bothers me. It does not seem like such a small step to go from 'decentralized' to 'cartel of relayers' to 'central bank and subordinate branches'. And the fact that this protocol is unavailable to US persons is interesting, though perhaps standard for the space right now.

The censorability of rollup relayers bothers me too. As you mentioned, the fact that US persons are blocked from the protocol shows this is a real problem.

I think this is a problem with technological solutions (perhaps anonymous and redundant relayers, or private rollup transactions so that the relayer does not know the content of transactions in a block but can produce a valid output state) that will be worked on in the next couple years now that the base technology (rollups) exists.

We need more precision in the scale and performance targets you’re seeking prior to discussing blockchain analogs.

It’s tough to respond when “serious financial application” or “revolutionary technology backing all finance” aren’t well-defined.

Sorry, thought my last paragraph covered that. Can any blockchain-backed tech handle the number of transactions per second that current major credit card/payment networks do?

One benchmark is 1,700 tps for VISA. https://phemex.com/blogs/what-is-transactions-per-second-tps

If not, why not? Especially after 10+ years of development and intense VC funding (as is the parent article's point).

Solana’s on the same order of magnitude as Visa. Some people dispute the exact figure, but it’s basically there: https://www.benzinga.com/amp/content/25031541

Next long term target is speed and energy efficiency of a Google search.

Edit: Avalanche says they’re around 4.5k tps: https://support.avax.network/en/articles/5325146-what-is-tra...

Fastpay can pretty much do infinite transactions/s
1,700 tps for VISA is a good benchmark. I recall VISA's technical capability is 10x or 100x that number,

Ethereum 1.0 can handle 30 transactions per second.

Part of the development dubbed "Ethereum 2.0" is focused on scaling the number of transactions via sharding. Each shard will be able to handle 2,400 tps. As more shards are deployed up to 64, Ethereum 2.0 will reach 160,000 tps.

I am interested in learning about how "shard chains" work.

I'm concerned after reading https://ethereum.org/en/eth2/shard-chains/ that sharding is aimed at letting individual dApps roll up transactions -- i.e. the individual app would be the effective shard key -- and thus that would introduce some notion of centralization into the system.

I imagine there are some good podcasts or articles discussing sharing in greater depth. I'm interested as well if anyone has suggestions!