67 comments

[ 6.8 ms ] story [ 134 ms ] thread
genius move from amazon. correct me if im wrong, this looks like a blockchain backend as a service?
But why are we celebrating centralization of systems that have "decentralization" as their biggest selling point?
It makes the numbers go up!
I'm not sure we're celebrating it. It's something to be noted. Some applications might find it useful, for running data analysis on the blockchain, but I don't foresee a majority of ETH validators running on this service. Best to avoid Amazon in general given their propensity for censorship.
How is one more vendor offering to run a blockchain vm centralization? It's just a hosting service.
amazon getting into crypto adds a ton of credibility from an outsider's perspective. even if you're not a fan of amazon, this should have a positive impact on crypto as a whole.
This is basically half of ankr’s business. They do back and block chain for large companies as well as start ups, and even PCs that want to stake their tokens.
Just did some quick back of the envelope calculations on profit from staking using this. At current price/rewards it would be just breakeven. Probably better suited to some corporate use case where you are more interested in avoiding fees rather than trying to make money from staking.
This is going to be big - blockchain and smart contracts as a service, that's the new "mobile apps"
Expensive though. Remember when consumers were tough on spending $0.99 for an app? Imagine them seeing $100 gas fees.
(comment deleted)
People pay $100 fees all the time in return for goods and services that provide some utility to them. The problem is "blockchains" don't seem to provide much utility to anybody at all.
The gas is paid in ETH, which is only backed by 30c each (even less if you count premine).

  julia> 100.00 tethers / (1670.27 tethers / $0.30 )
  0.017961167954881545
So in real terms the current gas fees are just 1.8c ;)

And yes, there are people and smart contracts who are today paying 1670.27 tethers to get control of 1 ETH. Even worse there are people who are paying 1670.27 real US dollars for the same, but it will not last for long.

Ethereum ICO funding created a huge incentive misalignment between investors and users. The ETH was crowdfunded as the "fuel for the worldwide computer", kind of AWS Credits, but in reality it was a stake/shares in an Ethereum mainnet network.

Why? What actual problem does it solve for the average user?
So now the ETH2 folks will have to coordinate with Amazon in getting that managed service migrated to PoS? How do we know that this wont be like ElasticSearch, in that Amazon runs its own custom ETH1 chain for its customers forever?

Another reason to doubt that ETH2 is actually going to be a thing sometime before the heat death of the universe.

Why would AWS have more clout than the whole rest of the Ethereum ecosystem?
They dont, however, they may end up becoming their own ethereum ecosystem.
AWS probably has more money than the rest of the Ethereum ecosystem? Anyway, even if they turn out not to have more clout it could still end up as an ecosystem split where some will migrate to ETH2 and some will stay on the AWS managed ETH1 version.
"Amazon Managed Blockchain ... automates Ethereum software upgrades"
It's conceivable that over half of all Ethereum nodes could be running on AWS in the near future. Looks like 41% right now: https://www.ethernodes.org/networkType/Hosting
And yet, 85% of geth nodes are running recent versions. The idea that AWS is going to force people into running old software is a conspiracy theory.
I went to try out a friends dapp based lottery pool. To transfer 500$ into the pool I had to convert my tokens then pay around 150$ in gas fees to get it in there, then another 150$ to get it out. So, while Im a believer in distributed systems, we got some work to do.
You're right, those fees aren't transferring nearly enough money to early adopters. It could probably be made to emit more CO2 along the way too.
Yeah gas fees are insane, more and more stuff need to happen on L2. and ETH2 can't come fast enough.
Optimistic rollups in the short-term, but I'm very hopeful for EVM zkRollups in the next 6 months: https://zksync.io/faq/intro.html
Loopring (https://exchange.loopring.io) is already here, also based on zk-rollups.
This is correct, but Loopring can only do basic trades. It can't run arbitrary Solidity code...yet. zkSync, that I linked above, is working on bringing Solidity and the Ethereum VM to zkRollups. They claim they can pull it off by this summer. Many DeFi projects, including Loopring, invested in zkSync as a result. I'm hoping they pull it off.
Perhaps they could use https://cheapeth.org/
will I be cancelled or praised in India if I launch pajeeth? feel like it could go either way
Oh you mean the one where the author did a huge premine for himself and bans anyone who dares mention that?
My favorite thing with L2 is how you can call almost anything an L2 and people believe it. You can just tell them to use an independent proof of stake network with permissionless validators like Polygon (formerly Matic), and everyone believes this is the "trustless growing L2" while Binance Smart Chain - which is also an independent proof of stake network with permissionless validators - is the "centralized" one.

Polygon has a separate Plasma network which is not what anybody uses for all their smart contract redeployments.

The gulf of knowledge just gets wider and wider, and more and more profitable.

Yeah a trusted 3rd party guide that calls out the security properties of the different L2 & other scaling approaches would be really good. I'm also concerned about how so many approaches rely on security though making a ton of on chain tx disputing any bad actor. Seems like a DoS vector
some people are even worried that the current fees will last just long enough for competitors to pull the rug under eth feet
I'm not sure about this. It is optimistic to see big actors getting into Ethereum but i don't think we should be happy about it. A "centralized" decentralized network running on Amazon servers? Are we learning from the past or are we repeating the same mistakes again?
I think it's a positive. If large institutions replace chunks of their legacy financial infrastructure with decentralized, lower-cost infrastructure running over large, distributed blockchains (Bitcoin, Ethereum), we would all be better off.
This isn't about running Ethereum on Amazon servers. They're not talking about mining or staking.

This is about making it easy to connect corporate networks to the public Ethereum chain. It's a full node, so it's talking directly to the p2p network instead of sending your transactions through some third-party provider like Infura, but it's not producing blocks.

Well, ETH prices didn’t seem to react to this noticeably.

Do you guys know what are some events that reliably move the needle?

Currently, ETH is in big trouble: delay on the POS (I don't think it'll be ready before 2022), new binary compatible chains that will allow everyone to move on if ETH2 fails.

Also the ETH futures recently introduced will initially push the price down, like they did for BTC.

Just to clarify that a bit, PoS is running in production right now with 100K validators and $5 billion staked, with four independent client implementations. But for now it's running in parallel with the old chain, which might not merge in until next year.

In the meantime, lots of layer-2 stuff is getting rolled out this year.

given amazon's historical strategy of attack against open source platforms and resulting trail of dead, this does not bode well for eth. i still expect the price to jump in anticipation
Not saying amazon's way of interacting with OSS has been great, but which ones have actually died?
Instead of reading this press release, please read this seminal article from 2001 rebutting it: https://nakamotoinstitute.org/trusted-third-parties/

“Amazon managed blockchain” is an amazing phrase. All three words are wrong. “Blockchain” is a distributed technology created specifically for censorship resistance. A “managed blockchain” should just be replaced with a database. Databases are faster & cheaper.

Meanwhile, Amazon is a company that has engaged in censorship many times over the years.

And that’s not even addressing the issues with ethereum.

The publisher censored Dr. Seuss by ceasing to publish it.

Bad habit to conflate one company's own choice as a statement on society and actual consensus.

The only way you would have done that is by having a social circle that manipulates your feelings about "cancel culture". You should clear your cookies if you are that easy of a mark.

I wasn't making a statement on society, I was pointing out that Amazon has and can censor. There are many examples, going back to banning wikileaks that resulted in them having to accept bitcoin in donations — and now they never need donations again.

Bitcoin routed around Amazon’s censorship.

But historically, when you are supporting censorship, you’re not on the right side of history.

Bitcoin is fixing this.

Ah okay, I had seen other discussions with similar writings, I stand corrected about your intended predilections
what the heck with dr seuss? get your culture war BS outta here.
I agree with most of this comment, and how "Amazon Managed Blockchain" is ridiculous. But it's not censorship, and it's certainly not Amazon doing the censorship, when Dr. Seuss's own estate decides not to publish six of his books.
You might want to get basic facts right if you're going to spend this much energy being angry about something.
Or do read the press release, and see that this is not a managed blockchain. It's a full node that connects to the public Ethereum network to send transactions, but it doesn't produce blocks.

I.e. it's an easy way to connect your corporate IT to Ethereum, without having to keep up with software updates or send transactions through some third-party API.

Not your node, not your blockchain. Already Ethereum is centralized because there is no independent network of nodes that can do a full sync from scratch. It depends on a company hosting its servers on amazon (separate from this service) to sync.

This just adds to the centralization.

A node hosted by Amazon cannot fulfill the most basic promise of a blockchain. Censorship resistance.

And that’s why you should have read Nicks article.

If they're not producing blocks, then this does not give Amazon any censorship capability. The only thing they could do is refuse to send their own customer's transaction to the network, which of course would result in losing that customer.

Even if they did produce blocks, they'd be just another miner among many unless they got to 51%.

BTC maxi FUD detected...

I have a dappnode running under my desk that is running a full-node (minus mining) and didn't require anything special to sync beyond a large-ish SSD.

If Amazon goes down tomorrow and takes Infura with it, I will continue to interact with the blockchain. So, tell me exactly what you are going on about?

With all do respect, "connect corporate IT to Ethereum" is the most depressing thing I've read all day. Enjoy socialism for the rich.
You seem to be very pro bitcoin, why is ethereum depressing and "socialism for the rich" at all, let alone compared to other crypto currencies?

(also the term is 'all due respect')

If you run your own hardware you also have to trust that it wasn't compromised by the vendor. At some point you just have to decide which level of trust to give each vendors and for some people, trusting that Amazon will operate their clusters without invading their privacy or accessing their data is a pretty good bet. If found out, it'd be the end of AWS's reputation.
This absolutely true and a good point. In fact you gave to audit the code you run too. You have to decide who to trust. So how fo we mitigate that?

Decentralization & Cryptograhpy

Cryptography allows that the person you are trusting is the only one who could have changed it. For this we can use code signatures. And of course you tun your local copy (distribution) which you can audit, and compile yourself. That makes it much harder.

You can analyze your hardware and once its left the factory it us harder for the maker to influence.

The point isn’t that you csn economically get rid of all trust, but that choosing to trust a third party is a security hole.

> “Blockchain” is a distributed technology created specifically for censorship resistance.

No it isn’t. Blockchain is a merkle chain where order is defined by a consensus process.

Your comment is the usual existential criticism that nearly every blockchain thread on HN is plagued by... you misunderstand the use case as well.

“A car is an internal combustion engine driving wheels via a gearbox” is not a refutation of the statement “cars are transportation technology”

Ironically you were part right: Empty attempts to score points like yoyrs are a bane on any discussion of bitcoin in HN.

Hopefully IBM can offer consulting service for implementing apps built on top of this
And I hope an Oracle database is part of their recommended solution.
Does anyone use AWS or other cloud providers to mine crypto? Can't imagine it's profitable today, but wondering if there was anybody leveraging that in the early days to mine without buying a bunch of equipment.