"Once the Liquid Network goes live, there is a monthly subscription fee."
The whole Blockstream strategy may be on fire. The main reason crypto is in this tribal mess stems from the bridges they were willing to burn to try a SaaS model. https://t.co/DxNviSVfkt
A federation of currently 15 companies in the Bitcoin trading space sign blocks and hold joint custody of the coins in the Liquid network.
So basically every block in the network needs to be signed by 11/15 of them and for people to withdraw money out of it, 11/15 have to sign off. This means that for money to get stolen, 11 of the 15 companies need to work collaborate together to make an invalid withdrawal. On the other hand, 5 of the 15 together can put the network to a halt (f.e. when they are taken down by a DoS attack).
The devices used by these federation members are high-grade security devices with no remote access, custom offline hardware module for signing, ... that only and only verify blocks and verify withdrawals.
The big conclusion is that primarily if you already hold your money on an exchange, it's a 100% upgrade in security because instead of your exchange to be hacked, 11 of them need to be hacked. That's why traders with their bitcoins on an exchange are our core target user group. However for other users it can also bring value with added confidentiality and faster block times.
there are multiple mis-attributions so it's significantly exaggerated. it also doesnt show other entities eg chain-code. also where does the data come from is it talking about IRC meetings? i'm not sure why it would be surprising that protocol capable developers would attend them. you'd be more worried if a company had no one participating.
but that's a false premise that's been debunked multiple times. lightning and liquid are scaling for bitcoin, but mostly they are different and better tradeoffs for specific use-cases. it's not engineeringwise that realistic to assume $1mil international transfers 0.1c microtips, $3 coffee, and censorship resistance needed use cases and high value long term cold storage all need the same tradeoffs. one or more of these use-cases will always be sub-optimal in a one-size fits all world view where everything goes on chain.
also blockstream the company holds bitcoin, and all employees are partly paid in locked in Bitcoin so stand to gain from bitcoin adoption which depends on scaling bitcoin.
plus blockstream is one of the three companies working on lightning which is far better for scaling some types of transactions than flood fill broadcast global permanent record. and lightning payments are faster, nearly instantly final, lower cost, and more private.
How is the 'trust' in this federated chain work exactly?
What I mean is, what is the risk for using an exchange with Liquid vs. the present custodial system? Does liquid make it potentially more difficult for a single exchange to steal, lie about holdings, etc.?
Is it more the case that you're trusting the 'partners' as a group, rather than a particular partner?
so think about it as liquid is a blockchain, though with >2/3 threshold signing in place of mining. the risk depends on how the asset are held. if you hold them yourself with a hardware wallet (cold store) then your risk is similar to cold stored bitcoin for liquid assets, because you can run your own fullnode and the assets are native.
for LBTC the story is a bit different, however there are hardware modules with the keys, automated boxes stored in exchanges operations centres (typically same location they store their hot wallet, so it will be hard to get physical access). the keys are stored in the hardware module and do not leave it. there are also other protections, the hardware module will only peg-out to exchanges who can use a redundant liquid fullnode to verify it is correct.
because you can transfer faster deposit and then trade is more plausible. and potentially trade with multisig trades it can provide more direct user control and longer term reduce exchange hot-wallet risk.
Yeah but the way they peg bitcoins to lbtc is federated, which, if my understanding is correct, means token issuance is essentially federated too. And federated means "not trustless." You're still trusting people with your money when you go to put it on Liquid or take it off -- it's just that you're trusting a group instead of an individual. I'd love to be corrected, though, as I've done very little research on this.
It would need a soft fork for a peg to be trustlessly made. However the only soft fork that so far is in test net right now is Drivechain and it kinda has issues due to the fact much of it depends on miners to peg it. However it being worked on now so we got to wait and see.
The federated side chains exist now to be used and test further until a soft fork peg is made.
I'm actually quite a fan of a federated sidechain. For a lot (and I daresay most) transactions it makes sense to trade-off security for convenience. Done right, I think a federated sidechain is actually quite a good choice for things like "buying coffee".
That said, I'm a bit confused about liquid. From the website it says:
Once the Liquid Network goes live, there is a monthly subscription fee.
Which is extremely disappointing unless it's only referring to the federated block signers. I feel a far better business model would be to monetize (the tiny) transaction fees, and push for mass adoption.
> I feel a far better business model would be to monetize (the tiny) transaction fees, and push for mass adoption.
It would plausibly be a better business, but it wouldn't be good for users who switched from Bitcoin. Liquid requires much more computing power to validate than Bitcoin does, and its blocks are signed by a fixed set of known participants rather than the open mining ecosystem.
Basically, Liquid is a clear improvement over any system that involves trusting individual exchanges, but Bitcoin itself is definitely not such a system.
I actually like the federated aspect (no reorg logic is amazing to work with) just saying that the system should be (at scale) profitable to run based on (small) network fees.
Despite putting in effort and asking direct unanswered questions to adam (see my previous posts), it's really not clear exactly how the whole system works or is supposed to work.
What does "exchange-level" participants mean?
As an "end-user" can I send and receive liquid directly, without paying fees (other than network fees) and without getting permission? And then by extension if I'm a business, and I want to accept and receive liquid (on behalf of my customers) can I do so without subscription fees? And by extension, what if my business is an exchange?
(More specific details to follow) 1minute blocks, likely around the same throughput as Bitcoin mainchain -- around 10x more blockspace but around 10x bigger transactions due to a lot more Zero Knowledge Proofs with ECDL based ZKPs for confidential transactions and confidential assets.
More federation members will be able to join later.
> likely around the same throughput as Bitcoin mainchain
How are transactions selected then though? Do they also pay fees? And as consequence, is there a fee market expected, similar to the actual Bitcoin? If it takes off, then "slow" confirmation times are also expected?
yes it's very bitcoin-like. fees are paid in LBTC. block-construction is based on highest fee-rate. the difference is most of these transactions are likely to be trade related larger transfers, so they won't push up bitcoin mainnet fees.
possibly bigger could be later. it has the similar tradeoff to bitcoin, it is important for businesses and individual power users and traders to be able to practically validate the chain with their own fullnode.
As Adam says, Bulletproofs will make the transactions smaller, though the verification cost per byte stays roughly the same so this wouldn't affect the 1Mb limit. But it would increase the maximum number of transactions per block. We have some more crypto tricks up our sleeves which are similar.
Less excitingly, but more practically, as the network operates we will learn more about how transactions and blocks are propagated (and have more time to optimize this aspect of the network). I expect that this will let us reduce the block interval from 1 minute.
we can make the transactions smaller with bullet-proofs which has some log-scaling benefits also. possibly bigger could be later. it has the similar tradeoff to bitcoin, it is important for businesses and individual power users and traders to be able to practically validate the chain with their own fullnode.
Eh. I had higher hopes that this would be permissively licensed open source and not federated but it's still a tool with a purpose. Pretty much federated ripple on top of bitcoin. There's clearly some demand for a product like that, but I was really hoping for trustless and open source sidechains.
Liquid is essentially a vendored elements. The code is identical except for naming elementsd to liquidd. So all you need to join the network is the liquid-specific chain parameters and a way to find peers.
These will be made public soon enough, a bit more patience!
As a user, no. I mean no one can prevent you from signing blocks, but no one will attach any value to your signature as you are not part of the federation.
The current (15-member) federation can add new federation members through a protocol specific for that use case.
What then stops other federation members voting in a new member without Blockstreams approval, or voting out Blockstream and taking over the network? Is it all just via legal contracts? What else do those legal contracts enforce, can you share a template contract?
I have never seen the actual contract. All the software is open source, if they want a federation without our coordination, they can. The thing we provide them is the coordination, secure signing devices and support.
If there was a FOSS full implementation node I was going to investigate ways a small business offering financial products might benefit from LBTC.
Maybe we could benefit by enabling high velocity bitcoin transactions between our start up service and existing exchanges, possibly allowing my employer to tap into the volume in those markets without on chain transactions and long confirmation waits for our users.
All in all I don't know what I would use it for because until I play with it a little I'm not entirely sure how it would be superior to RSK or Counterparty or even as you suggest Elements for use cases I might need.
There was mention of a FOSS mobile wallet coming in the future, but I'm unsure if a full node able to issue assets and the like is coming to desktop.
I don't think "federated Ripple" is the right way to put this.
All of the Bitcoin is pegged, you can't create your own, and as Adam also mentioned, all transactions are confidential transactions.
The chain has finality where as in Ripple, as we all know, the few nodes running the show can make up the rules as they go.
That community doesn't care about immutability.
It's absolutely very different from ripple, but ripples goal as far as I understand it is to increase liquidity and reduce settlement times between financial institutions. That seems very similar to liquids foremost use case.
The description of a federated ripple on top of bitcoin seems to me to be as broadly fair as any 5 word description can be. It's secured by a federation as opposed to a centralization and runs on pegged units of bitcoin.
Other than that it's a broader platform than ripple as it enables the issuance of assets and as you note has an immutable block structure and consensus rule set. It also includes confidential transactions via zero knowledge proofs.
It's not a perfect comparison but what 5 word description is? "Federated ripple on top of bitcoin" seems to be what it does, give or take some caveats.
We shouldn't describe anything in terms as a kind of Ripple, especially anything Bitcoin related. The project started by coopting an honest idea, added technical mumbo jumbo, an instamine, and some suits.
And yet an enormous population of technical illiterates still buy and use it. Comparing liquid to ripple is to say "this does the same thing but better in every way". You don't think that's a fair way to describe liquid to the people gobbling up centralized scamcoins doing the same job as liquid?
I don't think so. Ripple is better mentioned as a crypto-inspired Visa/MC. My point is not to create any kind of mental dependency on Ripple. That way it won't sucker more people in who might just be interested in tech like liquid that is truly based on Bitcoin.
That being said, I would be interested to hear user numbers on Ripple. https://twitter.com/kerooke/status/1050494653953323008 suggests it's not worth mentioning.
Sure, i.e. federated. Maybe federated is better than centralized like XRP, but it's a far far cry from decentralized. It's unfortunate that so many layer 2 building companies have for the moment eschewed the more difficult challenges of trustless sidechains for the more immediately profitable closed system federated sidechains.
Liquid is a public network however. FOSS smart-phone wallet coming soon. You'll also be able to run a full-node and audit the network (though much of it is encrypted due to the confidential transactions and confidential assets.)
Note for native assets the federation signers can only select the transactions, once they are final, which happens after 2 (1min interval) blocks, they are not able to be undone - as it is not probabilistic, reorgs cant go deeper. So a full-node can verify that.
FOSS tools and examples coming. A liquid asset is just a special transaction type, you pay the LBTC fee and create them. Whether some exchanges will list your asset is a different topic. They can exist unlisted as demo/test assets.
I work for a startup that would love to be investigating these tool sets right now - is there a full node software we could be playing around with? A smartphone full node is less than ideal for business uses. Does it cost any kind of licensing fee for businesses to use LBTC?
The use case is for trading, if the alternative is single exchange custody federated is strictly better.
Also as liquid transactions use confidential transactions and confidential assets, improved fungibility in part offsets the federated limitation.