Ross Ulbricht's avatar
Ross Ulbricht 8 months ago
Been reading up on the nostr protocol and comparing it to the DSP I wrote about in 2021. Many of the details seem to have been worked out brilliantly, but one thing that worries me is the reliance on volunteer nodes. It would be nice if there was a business model for nodes and node competition at the protocol level. As it is, I wonder if nostr can scale.

Replies (159)

Concordo com ele, pode ser algo que esteja freiando o avanço do Nostr como mercado. "Tenho lido sobre o protocolo nostr e comparado-o com o DSP sobre o qual escrevi em 2021. Muitos dos detalhes parecem ter sido brilhantemente elaborados, mas uma coisa que me preocupa é a dependência de nós voluntários. Seria ótimo se houvesse um modelo de negócios para nodes e competição entre nodes à nível de protocolo. Do jeito que está, me pergunto se o nostr pode escalar".
Ross Ulbricht's avatar Ross Ulbricht
Been reading up on the nostr protocol and comparing it to the DSP I wrote about in 2021. Many of the details seem to have been worked out brilliantly, but one thing that worries me is the reliance on volunteer nodes. It would be nice if there was a business model for nodes and node competition at the protocol level. As it is, I wonder if nostr can scale.
View quoted note →
The monetization isnt necessarily part of protocol, but it doesnt need to be imo
Ross Ulbricht's avatar
Ross Ulbricht 8 months ago
Yeah, I saw that there are paid relays out there. The market will find a way, but the less the users have to deal with the better. Nodes are simple/dumb by design, so designing a business model for them so they are all competing on the same metric (data availability) shouldn't be tough.
There are many business models that support relay operators that already exist, from directly paying operators, brands offering it as a value add or having a relay that supports a client/app that earns revenue indirectly As small as we are we already have over a thousand relays and could support millions of users
Two ways to business-ify relays : - Subscription - filters, algos, data retention - Ads - a free relay can serve up some digital vomit - don't like it, delete it from your list. I think it'll scale.
Several business models for relays exist, such as subscriptions. nostr.wine is arguably our most successful paid relay with several relay types in their service catalog. Another would be services rendered for relay access as a packed service offering, such as nostrplebs.com and others. Additionally, you don't have to run a large scale relay. Small community relays are the future thanks to NIP-65 outbix model. That's the future and many applications support it. Damus and Primal are the last major holdouts. Once they move over to this model, which they've previously mentioned doing so, the smaller community relay models will make more sense for everyone and decentralization goes up as a nice bonus.
The team @GitCitadel has had some success with starting to host a github replacement on Nostr. And, eventually, we will offer paid tiers for the #Alexandria project, an entirely new way to publish literally anything, with the intention to build a truly uncorruptable, hyperlinked knowledge base, and especially providing the means for breaking the back of the likes of the paywalled research journals, and eventually the whole publishing industry. All integrated with nostr's emphasis on community through interoperability. We would welcome you to reach out and ask anything. Our resident media mogul, @Silberengel is asleep (she's in Germany) and will be very busy tomorrow (as many of us will be, since it is Easter), but... I'm usually responsive if you @ me.
Keychat's avatar
Keychat 8 months ago
Keychat is a chat app, built on Bitcoin ecash, Nostr protocol and Signal protocol. Keychat is inspired by the postal system — stamps, post offices, letters. Keychat uses Bitcoin ecash as stamps and Nostr relays as post offices. Keychat uses Signal protocol to ensure message encryption security and meta-data privacy.
Karadenizli's avatar
Karadenizli 8 months ago
Why should note hosting be commoditized? Can it even be commoditized? Compare it to renting servers. There isn't a single market price to host a web app. Every host has its own features, pricing structure, free tier etc. Same with relays. Why standardize income source? The default might be to pay to have your notes hosted, but you could just as easily have a format where a relay invests in scouring and indexing all kinds of posts, and users pay to have a curated feed of the best posts of the day, or maybe even a personalized feed. You could see relays that charge for notes on demand you can't find elsewhere where you give the note ID and pay them to get the note back. I even imagine there will be relays that hold on to notes that people request to delete and will charge to see them like a paid wayback machine. Even in the most common case where relays charge to host your notes, they will differentiate on many factors such as $/storage, $/bandwidth, uptime, speed, hosting location, redundancy, censorship resistance etc. Hosting is not something that can be commoditized. One thing for certain though, and it's biggest difference between your concept is the ad funded relays. I don't believe this can ever gonna be a significant part of Nostr and the best thing about it. It will be a thing when there is a company that builds out a full stack with a client that works with their own relays but even then it's not necessary for any standard protocol to pay for it. The client side decides what the users see. It always has last say in that field. Therefore, any attempt to fund a relay with ads will be fought off by client side filters if it ever gains significant market share.
Funny u say this because I literally messaged someone yesterday and said, that everything you were thinking of back in 2021 worked out with Nostr Nostr can scale as is imo The good news is that anyone can build on it So if there is something you would like to try with relays …. A better way… then do it💟
Ross Ulbricht's avatar
Ross Ulbricht 8 months ago
This is what I wrote about back in 2021 if you want to take a look: https://rossulbricht.medium.com/decentralize-social-media-cc47dcfd4f99 The basic idea is that users would pay for content delivery, but it would happen under the hood at the protocol level and be super cheap and plentiful because of node competition (I called them "content servers" back then). Your average user wouldn't know or care about it, wouldn't have to shop around for private nodes or run their own.
jb55's avatar
jb55 _@jb55.com 8 months ago
this is like saying "the web won't work because web servers won't be sustainable without business models". yet tons of people run webservers for fun at a loss.
But who will run the relays? 😉
Ross Ulbricht's avatar Ross Ulbricht
Been reading up on the nostr protocol and comparing it to the DSP I wrote about in 2021. Many of the details seem to have been worked out brilliantly, but one thing that worries me is the reliance on volunteer nodes. It would be nice if there was a business model for nodes and node competition at the protocol level. As it is, I wonder if nostr can scale.
View quoted note →
That’s quite interesting, thanks for sharing. And your right we have seen a lot of open projects being centralized and controlled by single entities a lot.
GM. Yeah, we decided to tie specific relays from independent operators into our clients, as defaults. Users pay us and we pay the operators. If someone wants a custom instance of #Alexandria, our suppliers help them get their custom relay setup, and they then also pay the suppliers. That way we carry the cost, and they can focus on the administration and security. We're using relays from @cloud fodder and @semisol. And, of course, we encourage everyone to run a personal relay as a data store, like nostr-rs-relay or #Citrine 🍋 (for mobile). That way, they have their own archive and can use the functionality offline. It makes a lot of sense to store your #ebooks on your phone or tablet, for instance.
Yes, the #Alexandria client has the paid wss://theforest.nostr1.com (social) and the free wss://thecitadel.nostr1.com (publication/documentation) relays. The #GitRepublic project, which includes the upcoming publicly-viewable git server, also includes the private, paid wss://gitcitadel.nostr1.com relay. Each of the three relays are just the first of a set of relays, run by different professional operators, on different continents. Our income is currently over Geyserfund, zaps, and the paid relays. You pay us and we pay the relay operators. We also self-host an instance of the #Realy relay, now, which is important to us for some of the features (RESTful API), and to have a non-strfry backup. We're debating adding that as publicly-readable, to our overall user landscape, as it would help with the #epaper implementation and add more redundancy.
Nostr relays are highly decentralized because we have negentropy and outbox models. That means its easier to cover the running costs, for every relay that isn't especially large and overrun with users. You can't actually see most relays, as they're just little apps running on people's cell phones or laptops. I run the #Citrine 🍋 relay on my Android phone, and it has hundreds of books on it, takes up only 630 MB of space and doesn't noticeable increase my mobile bandwidth usage, since I sync it when in WiFi. Probably costs me a few euro-cents per month, and contains the notes that are relevant to me.
Told you they didn't read that post, they just welcomed you and zapped you! 😂 😂 Love you man, you are an inspiration!
Tony Acid 's avatar
Tony Acid 8 months ago
Cypherpunk Paolo Ardoino is building infrastructure that allows building decentralized apps that don't require servers to work ( ). Their first app using this tech is a messenger called keet.io It's not open source yet, but they plan to release the code to the public once fundaments are solid. If it works well, it will allow to build any type of decentralized apps on top of it and with proposed structure (no servers required) economically, imho, it's more reliable than nostr. image
Tony Acid 's avatar
Tony Acid 8 months ago
I guess this depends on how much of the fundaments they want to lay before public release.
Tony Acid 's avatar
Tony Acid 8 months ago
For example, the Keet UX for now is not very good, so I can't even recommend it to anyone except early tech adopters. So if I was them (hole punch team), I would make it bit better before releasing the code. Building stuff that don't require servers have a lot unique challenges, so it takes time. For now I'm just watching and cheering them. Idea is very good, but pending execution. There are also some privacy issues as well, that needs to be resolved. All communication is encrypted, but because its p2p, there is a suspicion that peers leak their IPs to each other. On the other hand, privacy on Nostr is total downer. Still a lot of work in this space needs to be done.
I understand, still I think a good way to improve code is to open up. If there are vulnerabilities, there's not much to hack at an alpha stage, so it would be a good trade off to open source. From what I heard, Paolo is doing a lot operationally, maybe he's busy.
Tony Acid 's avatar
Tony Acid 8 months ago
I just remember he said they plan open source it. I agree with you, sooner you do it, better, at least from security stand point. But I guess there are trade offs. Dunno. Observing the project for now...
BTCandT's avatar
BTCandT 8 months ago
Is it possible for everyone to run relays? Thought being, the more decentralized the better
If you're building a product on Nostr, chances are high that you will also be providing a relay for the network So the number of relays will increase along with the number of products
Lucas M's avatar
Lucas M 8 months ago
Good morning☀️🫡! Hope your enjoying Easter. Also, I have a question: do you recommend leaving these blank? image
The day you were freed, I posted the following note. And, speaking of business models, I'm still skeptical about the prominent role you outlined for advertisers. Is that still your view, or did the Nostr design show you other avenues? Thanks @jack for bringing @Ross Ulbricht to Nostr! nevent1qvzqqqqqqypzq086s94mfzf056lfjwk8920umwcgn002p3weqy0aypx32329lgkeqqstqgawzrqw5dztpcc8y78svpk7je8lsjre3zunfnzul02fhuyy23c5j638r
And yet, there are already over 1,000 active relays. Pretty wild. Some are paid, most are free. There are multiple commercial Nostr projects with an incentive to run relays. Primal runs two: one fully open, one for our premium users. We also run an open-source indexer and caching service. In my view, Nostr scales through a combination of the outbox model (a.k.a. gossip), caching layers, and eventually, p2p transfers between clients. Nostr doesn’t need to rely on a single transport protocol. Every signed event stands on its own. We’ll make many copies of signed events and find different ways to move them around.
I believe the best way is to make the purchase of relays very seamless within Nostr clients, either as an included feature for Premium access, or one-click rent your own. I will attempt to make that into my new Nostr client (Nostria). Perhaps the issue before has been that you purchase access to a larger shared relay, instead people should purchase their own relays. As long as Nostr clients implement the outbox-model nicely, it shouldn't be an issue discovering users and their content. I wrote about "Discovery Relays" here, that should solve it and then everyone can have their own smaller set of relays, not needing large popular relays: https://medium.com/@sondreb/discovery-relays-e2b0bd00feec
optionality is key. the freemium through ads model is clearly valued on the internet, but the ability for an individual to start with or upgrade to a different model would actually apply pressure on the freemium ad models to not be so harmful to the end users
Ross, once you understand nostr better you will find that "the protocol" is not even well defined. It's a loose collection of NIPs (on which we mostly agree, especially NIP-01). There is at least one NIP addressing your concern, see A business model for nodes already emerged spontaneously anyway, one of those is products subsidizing free relays. We should let free markets will do their thing.
Regarding over 1,000 active relays: how evenly distributed is the data across these relays? If I spin up a relay and nobody uses me, does my relay matter?
By nodes you mean relays? Here’s a business model: Personalized Webs of Trust relays. You’ll want to run one as an individual because it will keep track of who you trust and in what context. Running the relay yourself means you’re in full control of the algorithms that calculate your WoT scores. You’ll sync your relay with big relays like primal or damus but also with trusted friend relays which will make the big centralized ones redundant / unnecessary. The above product is what I’m building now. Calling it Brainstorm. 🧠 ⚡️ It uses strfry under the hood, arguably the powerhouse relay of nostr. I’m keeping negentropy activated so you can sync your Brainstorm relay with trusted friendly Brainstorms quickly and efficiently. Whatever content you care about goes into your relay. Personalized WoT scores are calculated using a variety of methods, including PageRank for noncontextual trust and the much more versatile GrapeRank 🍇 for contextual trust. You’ll be able to run your own Brainstorm. But as a business model, a service like @relaytools can host your Brainstorm for you. Imagine if things like Trending content and Recommended Follows are communicated to primal via your Brainstorm. Personalized to the max. No more global feeds!
Once a handful of people are using Brainstorm, you’ll have the option of syncing with them. If you don’t know who they are, sync with whatever relays that you know about, like primal or damus, and them switch to decentralized relays once your WoT tells you which ones exist and can be trusted. With Brainstorm, the initial baseline GrapeRank WoT scores are to sort the real users from the bots, nothing more complicated than that. Nothing contextual, doesn’t require all that much data. For now, I’m using follows, mutes and reports and those can sync up pretty quickly. Once the bots are filtered out, you can start to build a few contextual trust metrics, and one of the first will be trusted relays.
Yes you can use Primal Premium without applying our NIP-05 verified name. You can leave that blank in your profile or use another Nostr verification service.
@Ross Ulbricht immediately asks the most important question imo. Do we ultimately need "pay to post" incentives? Will this be solved by other emergent business models (as we're already starting to see from companies like @primal ). My intuition is that other business models will subsidize and keep the network sufficiently decentralized (ie enough relays storing notes). I also think that Nostr is more about the identity layer and ability to find peers or other communities via relay rendezvous points, at which point you can use whatever transport layer you want. But we'll see... Probably the biggest open question in my mind. Wdyt? View quoted note →
The relay data is valuable, can feed AIs, and more, there is no need to charge for subscription to relays. The relays will create their business models based on the data
I love that I can use my own relay, for my own notes. Similar to my own website, ran on my own hardware reliant only on my own internet connection.
Just feel incredibly happy to know you are free and able to share ideas with the rest of us. Happy Easter
Ross you probably right nostr will not scale. But at least its another decent option. But ya without billionaires donations nostr won't last but that's true for a lot of software. If a company see a benefit from nostr they will create foundations and support.
.. pretty quick to form an opinion :-)
Ross Ulbricht's avatar Ross Ulbricht
Been reading up on the nostr protocol and comparing it to the DSP I wrote about in 2021. Many of the details seem to have been worked out brilliantly, but one thing that worries me is the reliance on volunteer nodes. It would be nice if there was a business model for nodes and node competition at the protocol level. As it is, I wonder if nostr can scale.
View quoted note →
That’s good, because it doesn’t require arguments, the same way running your own node doesn’t require arguments You either do it for your own benefit, or you don’t.
Pfft. Wrong. If you don't pay, you are going to be doing more to usher in centralizing forces. Take responsibility for your usage and send attention/value via sats.
hasky's avatar
hasky 8 months ago
Yes true . They stole cable and sell it . Imagine that . But back to the main problem : what kind of business model that good for nostr …
:P's avatar
:P 8 months ago
You can run an instance of relay.tools with dozens of relays for less than 5000 sats a month.
I think at scale people will use one or two massive relays provided by a service, and then a couple volunteer nodes. Clients will likely provide a big relay, but other things like identifier, image hosting, search, etc., in exchange for a paid model or ads/boosted content. I think the clients will be doing a lot of heavy lifting in the future. We already see this happening with Primal.
bjorn's avatar
bjorn 8 months ago
So cool to see @Ross Ulbricht taking an interest in this little experiment
Ross Ulbricht's avatar Ross Ulbricht
Been reading up on the nostr protocol and comparing it to the DSP I wrote about in 2021. Many of the details seem to have been worked out brilliantly, but one thing that worries me is the reliance on volunteer nodes. It would be nice if there was a business model for nodes and node competition at the protocol level. As it is, I wonder if nostr can scale.
View quoted note →
Ross Ulbricht's avatar
Ross Ulbricht 8 months ago
Oh yes, I do. That's why I am interested in understanding it and looking for possible weaknesses. I want it to succeed long-term, and the best way to ensure that is to look ahead, iterate and improve.
We are very happy to have you and appreciate your efforts. If you ever have any questions be sure to ask the community as well. 🐶🫂
To be honest though when I first saw you had joined I thought you would probably attach your X to auto repost here then forget about us. Glad to be wrong.
Profit corrupts; volunteers endure. The protocol must enforce freedom so business models can’t. Scalability Fixes: Zaps for relays (tips to good operators). Client-side filters (reduce relay costs). Ephemeral events (auto-delete old data).
Ok. Not a dev, coder, particularly bright perspective here, but I would submit a gentle plea for space in a scale-less realm. A space where #nostr remains the cozy little aerie where the cool kids gather, away from the many-headed masses. One can dream.
Lucas M's avatar
Lucas M 8 months ago
The @GitCitadel team has been working on some very interesting projects that you might be interested in. I've been paying close attention and many others are beginning to, as well. 😎🫡
Laeserin 🇻🇦's avatar Laeserin 🇻🇦
Demo of #Alexandria's new Contact page, coming soon. The same #Markdown functionality will be available in our social feed (including long-form articles, calendar events, kind 1111 comments, images and video events, and #gnostr issues). No, that still isn't anywhere near as advanced as our reading viewer, based upon #Asciidoc.
View quoted note →
==================================== #7 ⚡ Most Zapped Last Week ==================================== Nostr’s Value4Value (V4V) model is all about plebs directly rewarding creators for the value they receive, no middlemen fees, no ads, just pure community-driven support using sats via the Bitcoin Lightning Network. Thanks to by @PABLOF7z for providing this data. Here are the top zapped posts from last week, showcasing creators who received the most engagement: 🔥 Top 3: Most Zaps 1. Zaps Received: 708 Sats Earned: 1M View quoted note → 2. Zaps Received: 155 Sats Earned: 42k View quoted note → 3. Zaps Received: 90 Sats Earned: 50k View quoted note → 💰 Top 3: Most Sats 1. Sats Earned: 1M Zaps Received: 708 View quoted note → 2. Sats Earned: 85k Zaps Received: 41 View quoted note → 3. Sats Earned: 50k Zaps Received: 90 View quoted note → #most-zapped_nostr_recap
bitcoin is the same: users have an incentive to run their own node and connect to a limited set of peers. its not the same as bitcoin in that: there is no global state. hence nostr network can scale infinitely but un-popular data wont spread that much, which is a feature and not a bug. #peercuration #opengroups
Default avatar
Internetbeard 8 months ago
Hey Ross welcome back! Im curious what you got planned?
Lucas M's avatar
Lucas M 8 months ago
I hadn't heard. Sorry, to hear that🫂. Honestly, I knew nothing about his impact on catholicism or anything related. Was he a good or bad pope, in your opinion?
hasky's avatar
hasky 8 months ago
Dunno 🤷‍♀️
The short answer is YES, #nostr scale. But the long answer doesn’t exist yet. We don’t know when he will scale or how he will do it without dying on the way.
Nostr can scale, but probably not in the traditional sense. I see many localized webs of trust on individual relays that will gossip notes to their desired destinations.
What I'm interested in with WoT relays is how you can gossip notes between them In a WoT model, you end up with a set of npubs you definitely want to fetch notes for. So the relay itself can subscribe to their notes from their outboxes. But if more people use WoT relays like this, then you could end up with many redundant subscriptions. So instead of a subscription, I was thinking of a gossip scheme where you ask a relay to notify you of just nevents for a given filter, and you can request the ones you're missing and vice versa. It's not exactly negentropy, but I guess it's kind of similar.
The strfry router command does exactly what you’re asking for, if I understand you correctly. Just use your WoT to decide which friendly and trustworthy relays you want to open streams with. Use a different filter with each relay if you wish, like if you trust relay1 for wikis, relay2 for zaps, relay3 for kind1, relay4 for profile data, etc. One of many reasons I’m happy with using strfry for Brainstorm.
My current version of Brainstorm runs a router to stream events and has several streams set up by default. I have a toggle button to edit strfry-router.config when your WoT is calculated (including your whitelist) and you’re ready to start streaming kind1 notes and other content.
Yea I think that the promise of nostr is a real potential for horizontal scaling (more relays) without needing vertical scaling (beefier relays). I think that worrying about the commercialization of the platform is a non-starter, because it's heading in a fundamentally different direction than Twitter is. Nostr will thrive when it's implemented hyper-locally.
LFG! keep poking holes! nostr needs your criticism and also your advocacy
The nice thing is that nostr is truly decentralized, so it doesn't really need to "scale". You can self-host or uncle jim, in which case volunteer nodes are fine. The ones that most need a business model are indexers like purplepag.es, or hubs that new/non technical users can use to easily get started like relay.damus.io, and those can likely be subsidized by businesses built on nostr. Then there are services like nostr.wine, which charge subscriptions. There's also been some experimentation in usage based pricing facilitated by ecash. So I think the problem isn't currently solved, but there are lots of leads to pull when it becomes more necessary. And it will, as spam of various kinds becomes more of a problem
Gregor's avatar
Gregor 8 months ago
I finally had the chance to look this over. I admit that I only skimmed over it, since my skepticism has no connection to the suggested corporate or distributed issue. Furthermore, my only personal attempt at public collaborative development has subjectively turned out entirely unfeasible and ineffective. In regard to the supposed dichotic alternative of corporate development, the current definition of a corporation, closely tied to the nation state model, looks entirely inadequate for the reality of global real time data networks. I might lack enough education to evaluate and articulate this more clearly, I feel entitled to at least describing conclusions from year long personal experiences. In conclusion, I do think that both ideas and implementations can and should speak for themselves, and current Nostr developments give me a definite sense of both flexible and durable properties, and a ground up construction. Pubky lacks that impression, though it did have comparably less time to grow. As said in another comment today, I personally just want decent basic functionality. I constantly see things do too much, and create far too much noise, including that elaborate treatise about a pretty tangential, and anachronistic issue. I would really like to see new collaborative structures that properly address attention economics in context of the internet, while avoiding the crutch of violence monopoly enforcement.
Gotcha. I think if you are still curious, you will find Pubky.app interesting, particularly in the context of the attention economics (i.e. end of doomscrolling and have the ability to become the algorithm and define what you see). We will be aiming to make the app public, in June. As for the rest, give it time, we are only getting started :).
Gregor's avatar
Gregor 8 months ago
I think artificial nothing
Esc's avatar
Esc 8 months ago
It’s a step in the right direction but still WIP I feel.
degenrocket's avatar
degenrocket 8 months ago
> It would be nice if there was a business model for nodes and node competition at the protocol level. No, that's a recipe for creating a closed ecosystem that limits freedoms and stifles innovation similar to having a strong federal government that sets all the rules like requiring all devices to have a USB-C port. Unlike Nostr or other decentralized social media solutions, Spasm is the future of social media because it's the only fully agnostic open ecosystem that doesn't limit developers or users.
Keychat's avatar
Keychat 8 months ago
Keychat has introduced the sats stamp concept, which is quite similar to what you’re talking about.👇 Keychat is a chat app, built on Bitcoin ecash, Nostr protocol and Signal/MLS protocol. Keychat is inspired by the postal system — stamps, post offices, letters. Keychat uses Bitcoin ecash as stamps and Nostr relays as post offices. Keychat uses Signal/MLS protocol to ensure message encryption security and meta-data privacy.
Keychat's avatar
Keychat 8 months ago
We hope that Nostr relays can evolve into post offices that sustain themselves by collecting ecash sats stamps.
Default avatar
Beraldo 8 months ago
Clients are doing this as Damus NWC.
You are 100 % right . We need to incentivize nodes some how. Zapping is a tool we have. All we need is the idea.