Thread

As a relay operator, this is the issue. No one wants to actually pay. @nostr.land is already working on these so called use cases “which paid relay operators are not incentivized to support” without needing ecash. No one wants to pay for relays or media hosting until it bites them in the ass. Many things, like the <$0.5/mo price figure listed on the post, are examples of people’s price expectations for a premium service. These prices are simply unsustainable for relay operators unless there is more demand (that is larger than the current usercount of Nostr).

Yup. And I think eventually we'll see primal, Damus, Yakihonne, etc clients that run their own relays eventually bake it into their premium plans. Primal already did this with nip05 and media.

People don’t want to pay. They also don’t want to see any ads of any form including privacy preserving ones. Then they complain about the crappy service. Then they send 5 paid relays’ worth of money to a grant org controlled by anti-FOSS people that then gets wasted on Pablo’s half baked prototypes.

You’re not wrong at all small scale. But we’re still at the beginning. Heard someone recently suggest that Nostr wins in the long run because it’s the cockroach that refuses to die. And even projects like signal and simplex run into funding issues to run the relays. I really do think micro payments has been a missing technical component, we’ve never practically been able to stream tiny amounts of money privately. For sure for this all to work Nostr needs to 100x in user count, if not more.

Could you elaborate on how you’re thinking about solving it without ecash?

Ecash payments don’t work. Relays being paid per connection or per event is not a sustainable model as the cost for the operatoe is mostly the same even if you post 50 times a day or 1. And $0.5/mo doesn’t work unless you have millions and millions of users that all pay *to a single service*. No one selling it for that cheap will provide you a quality service. You also can’t reach the required scale for economies of scale without people paying for relays *now*.

What do you mean? I paid for several relays and I am happy about the service. I know other people paid for it too and they also paid for media hosting. Lots of people are willing to pay.

x seems big because it's full of freeloaders, scammers and bots, nostr is not for freeloaders, it's organic and will win because it's not full of freeloaders. If you want to play, you must pay!

Cool But lots of people don't One of the small solutions, I guess is to lower the payment price, around 15k to 20k sats for per relay. (Not all relays priced like this but I do see some) That's too high atm imho Maybe lower them a lil bit? 🤔

Nostr.land does 5k sats/mo for an aggregator and 5 other types of relays in one. One time fees for relays as done by some big relays are ponzi schemes. New users subsidize the cost required to provide services to older users, as server costs are recurring, but payments are not.

it sounds all cool and all, until you realize the ecash is no different than topping up your account. (advance payment).

what account? There is no account with this scheme

ecash based models tie prices to nonsense metrics that push high activity users away while undercharging low activity users even though they use a similar amount of resources in the grand scale of things, which is not correlated to “activity” measured in events

Correct on subscription pay. Disagree on ads. This narrative is pushed by a fringe cult within Nostr. Ads fuel every platform that exists today and has existed for 50yrs. Ads will be a part of Nostr. But in a way that's: -Tolerable, not obnoxious -Private, not spyware -Respectful of users time/attention -Funds creators, relays, clients I'm working on bringing all this together in a way that works. Nostr diehards can hate me all they want, but the current model won't sustain the ecosystem. Responsible ads will.

XD

Stupid model, but that works almost everywhere. You're kind of describing support departments. The are entirely funded off added cost to the purchase price of the product, but more accurately, customers who paid for the product and didn't use support paid for other people to use support. The same goes for shipping in a slightly different context. We made a model of weighted averages where essentially a "tax" was added to the median products to subsidize the shipping cost on the larger products. So the more frequent purchases made up the difference for the more infrequent (heavier/long distance) purchases. obvi this happens everywhere, I just wanted to break down that this is not new and companies everywhere do something similar because selling anything almost anywhere is "unsustainable" in many ways.

Yeah but in this case, it’s a one time fee for a recurring expense. And the amount of money that comes in has to keep increasing over time semi-exponentially to be sustainable, or the costs need to be cut (you get worse service). This is worse than cross subsidization which is doable if managed right.

Which I see that you recognize here. This model works _realy_ well to make offering products sustainable. Distribution is the right word I suppose.

running a relay shouldn't cost $500k/mo tho

you have fees, taxes, customer support, the volume of data at that many users is huge, software devs etc. and then there needs to be money left to reinvest into the product

my point is that business always has two sides. if your service is economically efficient, seemingly trivial payments might add up. but i agree that there's a minimum necessary volume

currently 10k users throwing $0.5/mo would hardly fund 1 full time dev and servers including taxes and lost opportunities in other work areas

Event volume is also disconnected from real load. It just means you push out high-event-volume customers that may actually cost the same (less active time per day/data fetched) So it just makes the problem worse than it would have been with an “accurate” usage model.

But yes, the pricing is designed with the “average customer” as the target on Nostr.land. There is no good usage metric and users shouldn’t have to worry about quotas unless you are in the 99.9%p.

> Yeah but in this case, it’s a one time fee for a recurring expense Yeah it hides the actual cost of offering a service. Appealing to user's ideals, which in some cases is useful. As someone who prefers one-time costs over reoccurring costs, I also understand that unless you can accurately bill me, its not a true model. While it's nice to save money when purchasing bulk, I prefer up-front costs for a sense of "freedom" and "completeness". I still think it's possible to sell accurately priced products with longer time frames. The same distribution still occurs right, if I pay you up front, you have access to more capital, which can be more valuable compared to minimum short term payments. Therefore a tax is added to short term payment plans and/or a discount applied to initial payments/contracts.

I don’t mind ads if the advertisers are zapping me directly and subsidizing my Nostr experience as well as supporting the infrastructure.

The point of ads is it goes to the infrastructure. Enough to fund it and that’s it. If you want to get paid for viewing then you could use another microapp for that.

I use 1 paid relay and run 1 paid relay. Everyone who runs a start9 should run a relay and have the meetup figure out how to include it in dues or something. Relays absent a group or business are dumb imo.

A one time price in this case is one layer of abstraction too far. It is unpredictable how long a lifetime user will stay, while relay costs are easily translateable to a “average cost per user per month” with little variance.

do you need $5k/mo in hardware? yes, paying a dev salary has been exceedingly difficult for over a decade, but i wouldn't expect a relay to need full time support? dunno. i have 128 epyc cores, 1TB of RAM, and 20 TB of disk space for less than $5k up front 🤷

In software, where the recurring cost is maintenance (and can be bound to a few years per edition), it is much easier to model one time edition fees (which is a good model!) But in “cloud” the recurring costs are not time-bound and there is no product without recurring costs.

You don’t need $5k/mo just for HW. You need it for HW + dev combined. You’d need some level of redundancy and customer support as well, and to be able to dedicate enough time to it which could mean dropping a job.

Accurate and fair. I would agree lifetime is too far out. At least on the hardware side of things, it generally apears to be within about the 60-month lease window where prices are somewhat stable. When new hardware is released from Nvidia, or Dell, or HP, give it about 1-2 years for machines to ship, then cost of new products and derivative products go up, and old harware sells of in proportion to new hardware prices. Meaning new hardware has steadily increased in MSRP, and older hardware prices fall in proportion. So machines appear to be losing their value "slower".

perhaps it is all in the packaging and marketing. maybe that needs to be invested into. just a suggestion. I am paying tech subscription for 8 yr no problem for me.

Paying for attention is good. @Mandana is working on a variation of this. Biggest problem to overcome is spam/trust. Otherwise I build a botnet of fake npubs to watch your ads and run you dry, with zero value to the advertiser. So that feature needs to built after every other part of Nostr IMO.

I look at it this way: -Pay the infrastructure (relay/client).... Get a robust working platform. -Pay the creators.... Get user migration. -Pay the users.... Get retention. Priority goes in that order.

Yes that's the general idea of the #atttentionmarketplace we're building at @nextblock. Opt-in, set your price, watch and get paid. We charge a 20% service fee on the whole campaign,but also split that 50/50 that with referrals. So really only 10% to keep a premium experience running. And everyone gets paid.

Renting at least gives you the option to upgrade to newer hardware without much cost change, at an overall premium in prices. But it also means you do not need someone in the DC to do physical work. Anything more than a few years also reduces your flexibility, and increases the "lag" between cost changes if any happen => you passing them on => you getting paid

Yeah it’s nuanced. Sending locked ecash is fee-less and if the relay runs the mint (which they probably want to do because they can’t be rate limited or trust reliability).