Spark is basically a CBDC that is connected to Bitcoin - The service provider can see all your transactions - The service provider can steal your money and close your account - Fractional reserves are possible

Replies (72)

.'s avatar
. 5 months ago
Doesn't it also send a kyc packet ahead of the transaction?
.'s avatar
. 5 months ago
I swear I saw an interview with marcus where it explained the permissioned nature whereby a transaction wpuld be rejected before it even sent if the kyc was flagged by the receiver.
It’s trivial for the service provider to deny service. I do not know what mechanisms they have in place for this at present but it is definitely technically possible.
jb55's avatar
jb55 _@jb55.com 5 months ago
non-custodial basically doesn't exist for dust amount without utxo ownership. i don't see how you'll ever get around that
Yup, great point that is often overlooked. Wouldn’t call this scenario custodial though. Economical unviable onchain. Sorta like walking into Five Guys with only $5.
I don't doubt there is a use case for this trust profile, but why claim that something is non custodial? Unless they're playing with words and they want to state that *self* custodial is different from *non* custodial
What tradeoffs? Either you are using your own node or you are using someone else’s node. It gives you full control and increased privacy. It’s cheap, small, and quiet. I don’t see any reason not to be self-sovereign. You should be running a Bitaxe too.
spark/arkade self-custody is a gray area. if you have money in spark that you want to send or withdraw and the creepy centralized company is not cooperating, you can do a unilateral exit. this is very hard and potentially costs more than your entire balance. none of the consumer friendly spark enabled wallets have it as a built in feature. you must perform an arcane ritual with dev tools, and every single one of your VTXOs needs to be unrolled to L1 using layers of fancy and expensive taproot transactions. you'll hate it and think it's awful. there is an edge case where lightspark can collude with a person who previously sent you money in order to double spend it, by using keys corresponding to an older state of your VTXO that they pinky promised to delete. they have no way of proving that they cannot do this save for "we used an intel TEE" or some silly nonsense like that. these statechain protocols also implement centralized stablecoins. you can even see it in blitz wallet today. while they are floating around in the statechain they are designed to be freezable. maybe these are taproot assets that the L1 can potentially understand, but even if you wanted to try and exit them to L1 nobody uses them on L1 and they will be useless.
this is the only way you can scale lightning to large numbers of people. the only way to "fix" lightning right now is by making it centralized and adding trust assumptions. you can bitch all you want about spark but you should redirect your frustration to the lack of building blocks available on L1. premature ossification will just lead to things like spark. it is time to think about doing better. covenant opcodes are most likely the next building blocks to get added.
Yes. We try to be honest with our users and encourage them to upgrade to self-custody as it becomes economically viable. Try out the experimental version in the app today or see it fleshed out in full in v0.13 next month.
For those addicted to podcasts, here @Seth For Privacy explains things about #spark and their integration into @Cake Wallet [Ungovernable Misfits] Cake Adds Lightning ⚡ | FREEDOM TECH FRIDAY 32 #ungovernableMisfits via #PodcastAddict
It's a must trust, can't verify situation. You're trusting them to have deleted the setup keys (it's impossible to prove you've deleted something). If they haven't, they can unilaterally withdraw your funds too!
idsera's avatar
idsera 5 months ago
Leia isso: "Spark’s trust model is minimal and transient: 1-of-n operators must act honestly at the moment of transfer, and even if all fail, your pre-signed exit transaction lets you reclaim your Bitcoin on L1 unilaterally. No permission from Spark or anyone else." Fonte:
idsera's avatar
idsera 5 months ago
Vc confia somente no momento da transação, após a transação ocorrer, ninguém pode te roubar. (E mesmo que tentassem, é só com o detentedor anterior, não com todos os que tem saldos na Spark) E vc precisa confiar em somente um da federação ser honesto deletando a chave. E convenhamos, não tem incentivo nenhum para que algum participante da federação não siga a regra, imagina os três? Sobre a Spark ver todas as transações, isso acontece com a Phoenix e não vejo ninguém fazer drama contra a Phoenix.
idsera's avatar
idsera 5 months ago
Sobre reservas fracionárias, não sei de onde ele tirou isso.
idsera's avatar
idsera 5 months ago
Almost impossible. And will grow to more. What are the incentives to all parties cheat?
idsera's avatar
idsera 5 months ago
Esse povo não quer adoção do Bitcoin, ficam dando uma de puristas e não propõem nada melhor. Tudo tem tradeoffs.
idsera's avatar
idsera 5 months ago
I say the same about you.
Afaik you have the unilateral exist way always possible with Spark if they try to KYC shotgun ? If unilateral exist is possible then how is it really different from a LSP disabling your channel if you don't KYC ? Last question, wouldn't blind signatures on top or Ark\Spark be a reasonable trade-off for good UX and almost impossible denial of service attack ?
And with anything from above they are out of business. Right? Spark is x months old, it will get better.
the slow erosion is already visible. but most people still can't see the writing on the wall. this is all downstream consequence of perverse narratives winning in the bitcoin memetic space for a decade now.
Spark server code isn't public. You have to go through their company to become an operator. RE LSP privacy, if you have client-side payment routing (which ZEUS has) you don't know where your customers are making payments to and from. It's night and day.
The privacy side of the issue is valid for sure, but the trust model in terme'of custody is not so bad from what I understand, if there are enough Spark Operator involved at least. Any LSP could also try to cheat as they perfectly know that no is running a watch tower so... The most important issue I see is about privacy indeed, still but I can't unsee Spark as a good thing compared to full custody that many people used through WOS until now. I prefer to see the glass half full. But we can imagine informal Spark entities runned by federations "Liquid like", with blinded signatures on top of spark. That would be cashu on steroids actually. Best case scenario is obviously Ark + CTV + chaumian ecash.
Largely, but not entirely. Ark service providers have full visibility into all transactions within its Ark instance. They will always be able to refuse you service, but you can unilaterally exit if they do. The big self-custody issue is that without covenants, providers can collude with the previous VTXO owner to rug you. tl;dr signigicant improvements over Ark, but still fraught, especially without covenants
Yes, that's true, but realistically what percentage of payments will be in-round? I think 95%+ will be out of round due to the prevalence of Lightning and lack of interoperability between Ark providers.
Monero Dog's avatar
Monero Dog 5 months ago
I guess it depends. Ark is probably. more appropriate for near-real-time finality, and likely very well suited to trade between members of a community. Lighting will always prevail for real time, coffee-bar type transactions. Perhaps secure enclaves (see: #Lexe wallet) could be used to reduce Ark liquidy requirements; vtxo refresh could then be delegated.
There is a private 🔒 mode for bitcoin transactions, however not for stablecoins on top of spark. With the private mode that breez SDK has, the spark indexer cannot see it
Noah Fischer's avatar
Noah Fischer 5 months ago
Spark’s custodial model does reintroduce trusted third parties, but calling it a CBDC is hyperbolic—it’s more like a Lightning Services Provider with surveillance risks. Reminds me of how centralized choke points (like the Strait of Hormuz for meds) create systemic fragility when we rely on single points of control.