Oddbean new post about login | logout more generally, you can use someone else's curation as your own only to the extent you trust the curator. PoW has the benefits you don't have to trust anyone, which is I argue an important property if you take the PoV of service providers like relay operators.
Relay operators would much rather get paid than having to verify proofs of work.
verifying them is the same price as verifying at all, except for counting the zeroes at the front but honestly, EC signatures are not going to be very expensive for very long if someone makes them valuable and expects that somehow isn't going to invite a market in signature crunching just like it did for ... idk, all the shitcoins before proof of stake came around, and you ever heard of this thing called "nicehash" i get so tired of hearing about proof of work these days i wish you would all grow a fucking brain
wo wo calm down cowboi lmao. What I have in mind is to use npub_pow / total_pow. So, if a miner heavily attacks this system, it will increase the denominator. Unlike shitcoins that need to mantain global consensus and are "defended" by the total pow online at that moment, this considers the total pow over the whole history. So it's a strict improvement. Then by re-using SHA256 one can use ASICS for Bitcoin (I think) and offer these PoW endorsements I talked about here https://pippellia.com/pippellia/Social+Graph/Navigating+the+social+graph#PoW%20endorsement. So, you mine blocks with the ASIC, but when the price people are willing to pay for PoW endorsement is slightly higher, you go and mine that. This should result in having basically the same price for hashrate.
ok, i've been working with proof of work since 2018 and i have an opinion based on that, you may proceed as you wish with turning elliptic curve signatures into a new inflating monetary asset as you wish
and i've tried to explain this, unless there is a difficulty consensus you are fucked, whoever has the most compute at their disposal can drown your "spam protected" channels the only non-bitcoin cryptocurrency that has survived has been monero and that's because they keep changing the proof every other year to prevent a market in asics developing basically the whole rest of the PoW coin business just gave up because it's a game of whack a mole you can't change the economic dynamics of this game and the only way to prevent spam on nostr is paid relays!!!!!!!!!!!1
I agree paying a relay solves the problem of spam for that relay, but that's not necessarely applicable to every service. Some services aren't directly monetizable (not because it's impossible, but because people won't pay for that), and thus another spam prevention mechanism must be used. Also, many people have PoW keys (you included), so there is clearly demand for these things. My intuition is to use these and/or PoW endorsements for anti-impersonation and spam prevention. > unless there is a difficulty consensus you are fucked, whoever has the most compute at their disposal can drown your "spam protected" channels. yes, it can happen, but: a) that's not going to kill nostr, unlike these shitcoins you mention b) it's going to cost the attacker a lot of money, and the attack will be public, so people will stop using this spam prevention mechanism and the attacker would have accomplished very little
if it's not monetizable, it's ded we already see this unfolding in real time and yes, when there is an economic incentive to spam, the PoW scheme is ded don't mistake ideals for practical reality ideals gives you shitcoins and fiat, reality gives you bitcoin i'm only even responding to this because in your reply and from what i know of you, you are capable of understanding the game theory and economics, otherwise it would be cold silence, i've had this disagreement with several others about PoW and they just did not comprehend what ded means
ded: dedicated aka: deferred enforcement departure
(o_O)
nope, it means finished, over, done, no more use, life over, end
i need a new browser & fuck Websters
https://later.com/social-media-glossary/ded/#:~:text=%22Ded%22%20is%20an%20internet%20slang,of%20order%2C%20or%20extremely%20tired. it's a meme on the internet, you see
> turning elliptic curve signatures into a new inflating monetary asset
this is not what I am saying; What I am saying is mining over a nostr event.
the event:
{ target: npub1xxxxxxxxxxxxxxxxxxxxxxxx
nonce: 59h998543...
}
If you sha256 this event, the result starts with a certain number of zeros. ah yeah, i forgot that was what it meant there's a lot of very fast SHA256 mining hardware that can be applied to this, so what i said about it being ded multiply that by two orders of magnitude sure, most current generation miners can't be persuaded to hash anything other than 80 bytes but there is still pretty fast ones that can be reprogrammed, you have to implement a sponge function either way and the limitation is just a question of how you write the firmware, mostly
genuine question: why did you mine your npub so that it ends with 'mleku'?
i switched to a milk-only diet for about 5 months starting in december last year, due to having no money due to suddenly learning my sponsorship was over as i just had landed in madeira, and mleko is the word for milk in yugoslavian languages, so i put the "u" on the end to reflect the serbo-croatian "position" case (at/on) and i built a key miner, yes, way back then, i've updated it a lot... this is the original that i used to make my current key, took 4 days: https://github.com/mleku/vainstr this is the updated version, which optionally (preferably) uses the bitcoin-core/secp256k1 CGO library and it can mine a mleku key in about 3-7 minutes https://github.com/mleku/nodl/tree/master/cmd/vainstr i have since quit the milk diet because it was too much lactose and i got fat, now i mainly eat hamburger patties and other forms of beef, and i've started using up my stash of psillium husk because it helps me digest fats easier, and my energy levels have gone up drastically in the last few days since starting on that
oh, thanks for sharing this interesting story, and sharing the code I used. However, I was more interested in why did you mind that key instead of using a random one?
it was a deal i made with myself to not go online until i mined a key haha
okay, so just for fun. I thought you did it as an anti-impersonation mechanism
well, at 4 days mining it was kinda, but now it's under an hour, i'd need to make a longer name next time... maybe i should re-adopt elfspice :)
also that's why i'm like "hmmm not sure this PoW thing on keys is gonna do you much good" when i can get 40 zero bits at the front of a key in minutes...
What if I take an ASIC, make it mine, and people can pay 10$ to me to have these PoW keys or even better these PoW endorsements (so you can use the same old key), in order to get a "golden badge" displayed next to their name on clients, as a form of anti-impersonation. Like, I would pay for it, and other would as well I think. And if the market for anti-impersonation service gets big enough, it can actually work at scale.
golden rule of PoW: if it enables income via spam, it will be adopted by spammers
also, then you still have to pay to not be flooded by spam, when you could just make relays paid
💯 full circle
all forms of spam prevention boil down to paid access, really bitcoin uses proof of work, and uniquely can use it because it is money, and there is only one money, and the vast graveyards of proof of work shitcoins stands testament to the there can be only one problem proof of work is to spam-prevent the PROVIDERS of the bitcoin service the users have to pay transaction fees, which is the spam prevention so the same applies to nostr, except that we don't have a consensus so there is no competition for that, we just have a user spam problem so, we just need transaction fees, and because it's a relay, and it's not just one-off service, you switch to subscription model, but the problem is you need auth to do paid subscription model!
Agreed! Except for the subscription model. Making content publication and micro-payment the exact same action is the way to go.
why is subscription model not? it has a lower cognitive burden, and requires the use of rate limiters when the user exceeds their service level you would not make even lightning transactions as often as you like and reply to notes
at this point in time subscription model makes sense because most LN payments are going to be custodial and the friction level is going to be high, i have to click and wait to do zaps as it is, you want me to do that for every like and reply? if it isn't in the protocol yet, then don't make unrealistic expectations about an immediately paid transaction model without recognising that this simply is impractical and has not only a high cognitive burden but a much higher message cycle cost
1. Subscriptions do not have a lower cognitive burden in highly competitive, open markets. They're nightmare to handle for the subscribers. There's a reason people don't have subscriptions to every bar they have some drinks in. 2. That's why eCash is inevitable
i have like 5 running subscriptions right now, youtube, vodafone, protonmail, i forget all of them even cognitive burden so high i forget all of them more code less talk
That's the point of cash. You pay and forget about it.
yes, they would, but not every service is *directly* monetizable.
Can you give an example of such a service?
thanks for the zap mate. For example, Damus relay is free to use because Damus the company monetize in a different way. But they still need some DDoS prevention, and here is where PoW keys/ endorsements are useful.
That sounds like Damus leaving sats on the table. What am I not getting?
> That sounds like Damus leaving sats on the table. idk, maybe. Not every service is directly monetizable I think, even in a bitcoinized world. Plenty of examples from non-bitcoin product and companies.
Every service is directly monetizable. When people choose to not monetize that is mostly because of trust/reputation/generosity. Like babysitting on the neighbor's kid, no problem. If you then decide to accept anyone's kid and think the fact that this kid already lived for XX days is gonna solve your problem, well... 🙃 I think you better just have a price list.
Agree with Pip, most relays will be paid for by monetizing other related services
Any examples?
Barely anything is monetized these days so I don't have actual great examples, but Damus making money with premium subscriptions can subsidize their relay operations. Similar for zap.store
That kind of business model doesn't work in high competition markets: nostr:naddr1qvzqqqr4gupzp22rfmsktmgpk2rtan7zwu00zuzax5maq5dnsu5g3xxvqr2u3pd7qqghxatzwd3hy6tsw35k7mn5v4shyucfw4z8l There aren't many devs here that have experience with monetization in very open, competitive markets. Most either have zero entrepreneurial background and/or come from some kind of SaaS model. Damus' model sounds like they're starting a bar: - where the beer is free, but it's a mix of all kinds of beer (of which 90% is straight up ~~piss~~ Heineken) - where you can stay for max 3 hours - where some people pay a monthly fee for staying longer and access to some backroom This, in a street where you have millions of other bars: - that all cater to their own very specific beer-lover audiences - that are paid, but beat anything else in terms of price-quality - that can change their pricing any second to adapt to the volatile market forces in the industry - where customers are treated as part of the value add of the bar
Love your metaphors 😅 I agree with you. Still relays won't necessarily be directly monetized. We'll see. We need to learn from high volume low margin industries like -I think- online travel agencies (kayak etc) where data is sort of open, volume high and competition fierce
Yup, low margin stuff will exist It's just not where you start Those commodity business models are built on top of the flights that all are monetized directly. And like you say, they only work be sure of the volume, that exists because of the lower level business models.
can't find your linked long-form
I love this analogy and I think shows a dichotomy in the nostr world. One side is trying to replicate Twitter where clients are aggregating feeds from many relays and making it feel more like a global public square. The other is thinking about smaller, niche communities with a barrier to entry of some sort (more akin to the early internet with forums). I think both can exist but I personally want to build for the latter and I think the latter is easier to monetize.