#FreePavel
Butβ¦
It is foolish how he thought Telegram could indefinitely operate a network without default E2E encryption and not play ball with regimes.
I see the same sort of presumption from Nostr devs/proponents ignoring the core architectural problem of needing relays.
Probably an unpopular opinion but it has to be said. Relays will be the death of Nostr but itβs not too late to fix it. The protocol is still in its infancy.
Login to reply
Replies (113)
What do you suggest instead of having relays?
I could add that a single private key is not how I would design a free speech protocol.
If it were up to me I would use a set of at least two private keys, one for regular Nostr use and one for admin control of the identity. If the regular key is compromized, the admin key could be used to regain control over the content.
I hope Nostr will become the free speech protocol we need.
Can you explain why relays are in your opinion an architectural problem for Nostr?
What alternatives do you see? For example: In the Ham Radio world, we have multiple redundant frequency- bands and modulation methods to communicate in case of a grid-down scenario. #hamradio
How are relays different from nodes ? Bitcoin has nodes.
"Probably an unpopular opinion but it has to be said. Relays will be the death of Nostr but itβs not too late to fix it. The protocol is still in its infancy."
β
How they are discovered (btc nodes uses dns seeds not only hard coded list), the number of nodes and how big they are for example?
That is definitely an unpopular opinion! I'm new here and no tech, but relays don't make sense in the same sentence as censorship resistance. There is every incentive to support a relay operator, but how can they deliver on their promise if they are targeted? They aren't connected to each other, right? Like a spoke without a rim on the outside holding everything together. A bit floppy!
Also struggling with the notion that relay owners can delete notes if requested. What's stopping them acting on their own initiative.
And I see in habla news an edit button option after publishing. How does that fit with the immutable content line?
I'm sure the greater minds of Nostr have thought of these things and, I'm all in, but if there are checks and balances in the protocol, the head of Nostr marketing dept might want to step up and point me to a 5 year old's explainer.
From my energetic forensics points of view attuning to my body's intuitive gnosis (because my body knows before I do), and to make it a bit etherial it's interesting that both times I happened to find myself in a relay panel discussion my body was literally convulsing. I was feeling so uncomfortable and in pain and feeling like I wanna flee. Which I believe I eventually did.
So I'd say it's probably something here to pay close attention to.
nostr:nevent1qqs0u0s8y333ywuef0vrmfpr6thrmj78rxjymr3rh95xpffx50qp7tqppemhxue69uhkummn9ekx7mp0qgsvnvvlln2ru6jlyweayugxecv7ftfdlzd6zqca63sh7x6ercggjegrqsqqqqqp3953su
It totally should be. But I think Nostr devs donβt see the issue.
I was corrected recently on that same assertion of "P2P just like bitcoin". Technically it's peer-to-node and miner.
Is that observation semantic nitpicking or is bitcoin more than P2P?
In principle, unless we have central directories, relays that are run decentrally by anyone with an interest in securing the peer to peer network , should be good
No ? I don't know the exact functionality of these relays. That functionality may be the problem.
Plebs running a bitcoin node could be incentivized to also run a Nostr relay with the same hardware? then would be like tens of thousands of relays around the world.. and with the same equipment we would have access to freedom money and freedom speech transmitting all around the world
Id say its semantic nitpicking (which can be a good thing) and that it also depends on what ur trying to understand about it. Bitcoin is much more than p2p though, because there is a social layer to it, incentive models, etc.
How long you have the phone off during its life time? 0,1%?
When I tried to solve the βcontent storage and propagation issueβ I found that it would be possible to solve it using an inventive model, where the content itself is attributed a value (no shitcoinery). Basically the network nodes would cluster in #opengroups https://github.com/baumbit/opengroup and what we think of today as relays, would be transformed into network cluster bridges.
nostr:nprofile1qqsr9cvzwc652r4m83d86ykplrnm9dg5gwdvzzn8ameanlvut35wy3gpz3mhxue69uhhyetvv9ujuerpd46hxtnfduq3qamnwvaz7tmwdaehgu3wwa5kuegpp4mhxue69uhkummn9ekx7mqcu9929 is on it. π€
And how would Nostr need improvement regarding this? Also need of E2E?
Yeah notedeck was built like this from the ground up, and damus ios is in the process of switching to it.
p2p is still hard. Iβm looking into local sync using negentropy ( nostr:nevent1qqst670lhr7lmttdl0nta8l2s6d37yk4vrhulnddtyftcak3vgyt07cpzpmhxue69uhkummnw3ezuamfdejszxrhwden5te0wajkccm0d4jjumn0wd68ytnhd9hx2qg5waehxw309aex2mrp0yhxgctdw4eju6t0qyxhwumn8ghj7mn0wvhxcmmv4duvv6 ) but relays aint going away
nostr:npub1g53mukxnjkcmr94fhryzkqutdz2ukq4ks0gvy5af25rgmwsl4ngq43drvk has been looking into webrtc p2p stuff. Still early days
Isnβt that exactly how nostr works though? Relays are like nodes, notes are like transactions in the mempool?
As long as there are enough nodes/relays itβs difficult/impossible to shutdown the whole network. Right?
If Nostr doesn't improve its resilience there will be a fork at some time.
The price of using Bitcoin (danger to persobal freedom) instead of Monero gets higher every year.
You canβt really βforkβ nostr. Not like you can fork bitcoin.
I guess you could say there are lots of βforksβ already, with different people experimenting with different NIPs. But guess what? Theyβre still (mostly) interoperable. Unlike something like bluesky, which is more like bitcoin. Change one tiny detail and you have to build your userbase from scratch. No interoperability whatsoever.
This is one of nostrβs greatest strengths.
are economic incentives the only ones we care about?
I kind of like the idea of conceptually distinguishing between clients and relays.
a client is an application that connects to relays to pull/push notes
a relay is a server that accepts connections and processes notes in the specified schema
those are abstract concepts. if an application implements both interfaces it's capable of P2P.
if you want to own your data, you need ro run a relay. isn't that where the incentives lie? the economic incentives are created by people who don't want to run their own relays.
I think the question then is: can only P2P exist? is there no room for something in between centralized platforms and P2P only?
Probably why keet and pear protocol is on the right track.
Interesting
nostr:nevent1qqs0u0s8y333ywuef0vrmfpr6thrmj78rxjymr3rh95xpffx50qp7tqppemhxue69uhkummn9ekx7mp0qgsvnvvlln2ru6jlyweayugxecv7ftfdlzd6zqca63sh7x6ercggjegrqsqqqqqp3953su
Isn't that what #citrine does? Puts a node in your phone?
Nostr's whole point is that it isn't P2P. Relays are fundamental to that. If your solution is P2P, it's not nostr.
Ideally, relays should not see any plaintext unencrypted content that would make them liable, but idk how that would work.
But yes, currently relays are a liability nightmare.
My first thought as well .. βdonβt we already have relays on phonesβ
Yes, 100% Hole punch seem to be doing a fantastic job.
Don't get me wrong, Nostr are also doing a fantastic job, it's the halfway house solution.
Not sure what the incentives are to running a relay, can anyone help me understand.
Very interesting. I just read all your GitHub readmes.
How do you do networking and p2p?
Data has to be stored somewhere that's retrievable. What do you suggest?
π―
Can relays be replaced by something based on a mesh of anonymous nodes that don't replicate everything but dispatch notes on enough nodes to ensure reliability?
A node connects to peers that it discovers. The peers have to compete for the available bandwidth. Those peers who offers the best content gets more bandwidth. Before downloading content, your node download scores for the content. Content is always identified using its hash. An average score is derived based on your own prefences (key words, WoT aka #opengroups, etc) and a download priority queue is created. Basically put, #treebit is not(!) a gossip protocol, not a distributed hashtable either. Its 100% p2p where your node only store the content you value (which is the content you want and content your node think your peers want). The whole design is based on incentives. I dare to imagine that this is might be a novel p2p protocol, because I never heard of it before (though that could be simple ignorance on my part).
The node itself can be a hidden onion server (code on repo defaults to TOR), but data can really be transferred using what ever channel is available. You can even send content (nostr notes here) using USB-stick and someone else can publish it (same as Nostr).
I also invented βarchive nodesβ and βbridge nodesβ which is basically what relays are in Nostr. This implies that indeed, we could implement #treebit on Nostr without actually making any big changes to Nostr.
t-y 4 kicking this thread off Samson!
We just have to keep fighting centralization in any way we can.
Also if you understand how hole-punch / pears works then youβd realise itβs not an issue.
What's the problem if anyone can run a relay? They can't censor all of them...
The nitpicking was prompted by a Nostrasia interview with Stewart Mackenzie. Apologies if I get the lingo wrapped around my neck, but I think he said that unless you can show him that the path is from private key to private key, don't even think about calling the network censorship resistant. And he was very critical about IP addresses too. Probably best to listen to him! His benchmark is a state attack, rather than keeping out Johnny Rotten Hacker. Not a bad height to set for a bar, though!
Does your design allow nodes to maintain separate contextual channels such that "I get content from a b and c for [topic X]" and "I get content from m n and p for [topic Z]"?
In my opinion, any decentralized content/trust system must allow nodes to highly trust a given peer in one context, but strongly distrust the SAME peer in another context.
[PΓ©tition] LibΓ©rez Pavel Durov : Dernier Rempart Contre la Censure NumΓ©rique - ProtΓ©geons Notre LibertΓ© en Ligne ! https://www.leslignesbougent.org/petitions/liberez-pavel-durov-dernier-rempart-contre-la-censure-numerique-protegeons-notre-liberte-en-ligne-18862/
need e2e private communities
Hey totally can you code this please and sign me up when you're done thanks.
clients with relays included talking over bluetooth
Keet could fix this but it keeps crashing and runs like a dog
Yea, but β¦
I agree that the currently altruistic relays aren't sustainable, they won't be able to store all the data for everyone for free with growing userbase. However I don't think nostr is doomed. But I think the end game has to be some kind of payed model if you wan't all your social history to be stored. Or all users storing their own data, acting as "mini relays", though I don't know if this is technically feasible.
What do you think?
nostr:note1lclqwfrrzgaejj7c8kjz85hw8h9uwxdyfk8z8wtgvzjjdg7qrukqs2dydg
the sooner we realize discovery is worthless, the better. you discover people in the realworld, then connect permissionless over nostr to strengthen that community! discovery is for influencing and schilling shit. i aint here for either of that. i want curated private communities i find or start myself, and the ability to sign my own messages.
nostr:nevent1qqs0u0s8y333ywuef0vrmfpr6thrmj78rxjymr3rh95xpffx50qp7tqppemhxue69uhkummn9ekx7mp0qgsvnvvlln2ru6jlyweayugxecv7ftfdlzd6zqca63sh7x6ercggjegrqsqqqqqp3953su
I agree with your opinion! Very important observation.
And to answer your question: Yes it perfectly suited for the kind of curation you mentioned. I went furtherβ¦ while my proof-of-concept has a suggested default content filtering; the hope is that users would design their own filtering algos using the basic primitives: #opengroups #peercuration #hashscore #tags and I left it open so that more such primitives could be invented and integrated. Instead of a βglobal content consensusβ protocol I made a βlocal normativeβ protocol model. Not opinionated. Designed to be re-redesigned. #treebit could have bridges to other networks such as Nostr.
Why not have everyone hold a copy of some arbitrarily old section of the history of posts (in their client, as a separate program on their device, whatever) with a choice to hold the entire history.
Then, you connect to some peers, add a post by communicating it with them, and they communicate with others and so on, adding it to their local copy, until everyone has the message? Essentially, everyone is a relay.
You could broadcast to some sunset of your peers and then wait for another subset to show the post to ensure that your message proliferated the network.
Obviously, this is extremely similar to how it works with cryptocurrencies, but the fact that there isn't a requirement that everyone must agree on what the chain looks like should make this process much simpler. You can just sort posts by timestamp and if some people don't have it, then so what. That's the trade-off for true decentralization.
The bigger the community the less e2ee matters as the likelihood of compromise increases exponentially. But we will still have it cause why not?
Explain it to me like Iβm a plebβ¦
Why is it a βproblem needing relaysβ? And whatβs your suggested solution?
Simple X + StartOS βself-sovereignβ server?
How else do you make peace with regimes tho? Lol
MLS > Mastodittoβ¦
It can only be P2P. Any weak points will be attacked.
Relays make Nostr centralized.
This is like saying websites make the web centralized. They are only centralized if many people use the same ones. In a larger nostr web there will be tens of thousands of relays or more.
Need to make every client a relay first. Relays are a risk to the protocol.
Anyone can does not mean anyone does.
Memos
Youβd store it yourself if every client is also a relay. You could mirror content for others too.
There are no incentives to run a relay other than wanting to be helpful.
Or routers make network centralized ! Somewhat ..
I wasnβt comparing it 1:1 in terms of encryption. I was just highlighting each has its own critical flaw.
And thatβs why Umbrelβs option to make it your own relay is very interesting.
Everyone who has their own bitcoin node should also do this. It will only store your own data, but thatβs a start. We can build from there.
Perhaps true short term rn, but it can change..
for sure so far havent found the right product/mkt fit to help people find and use paid relays easily. But i think this is the primary incentive over time. I sure hope so
Either Nostr evolves to have every client being a relay, or it will fail. The risk is the same as that of LSPs. Damus itself should definitely not be running a relay.
not sure its tennible having every mobile client serving data for others..
Its great as an idea but not sure about in practice
Yeah I agree with that. We have incentive to run one to ensure reliability for our users, and our subscription pays for it.
Apparently the largest legal threat is now βhosted encrypted notesβ (wtf?). I never thought it would be a big deal since we do not host any media.
Yeah the reason why nostr works is because its not a p2p protocol
I started out with 9 relays when I created my profile here. 4 are dead now and Nostr is a lot bigger. I donβt think itβs fair to say that there will be more relays as it grows and not have a clear incentive you can point to for why people will suddenly run more.
I think all the recent events with Telegram and Samourai should be a clear indicator of where things are going.
Having many relays doesnβt help if they all have nice URLs and the operators can be easily identified.
I have my own private relay with umbrel. Works great
π―
Many shortfalls plenty of time. Biuld and educate. We must not fail.
Pear is using censorship/moderation in chats. Its not public. Its both a feature and a bug. Different use-cases. I love both Nostr and Pear, but they cant replace eachother.
How are relays different from bitcoin nodes? Surely the more we have the more decentralised it is?
"no clear incentive" argument doesn't cut it, never did.
There is 100m self-hosted sites on the web, and owners _pay_ to host every one of them. Their incentive is their business/hobby/etc. On Nostr, creators/businesses have the same incentive to run their own relays and to pay for good relays to have their notes delivered to followers. There would be at least 10m relays if Nostr got widely adopted, not counting all those phone relays serving as local caches.
Creators pay for web hosting. Creators will pay for nostr relays/media-hosting/etc.
I know free social platforms made this a bit non-obvious, but please look around the web. Hundred million people and businesses pay to have their own online space. Nothing new here.
You may indeed be using fewer relays now, but does the death of those four mean that others weren't created? I've seen a lot more people posting about running their own relays than I did earlier on. I'm even investigating how to run my own relay. What you see in your client relay list may not necessarily indicate the overall relay numbers. Am I missing something here?
No. People do.
I hope you are right. People try to conflate nostr and bitcoin growth and I cannot see it at all. I always look for the incentives, every aspect of Bitcoin provides positive incentives to human greed. I do not see those incentives for running a relay. The profit/risk ratio does not seem (to me) enough of a driving force for an ever growing number of relays. I hope I am wrong.
Disagree.
That's ok
Using this standard, what is decentralized? I'm not convinced that either end of the spectrum actually exists in reality or even can exist. You start getting into philosophy the deeper you dig, but even the most centralized systems aren't entirely centralized. Consider Twitter. Yes, that's pretty centralized, but Twitter doesn't own the ISPs required for it to exist, etc. Something like BitTorrent is pretty decentralized, but the peers are still owned by people who can therefore have some control over it. Someone still has to seed things on controlled hardware for it to matter. And the networks for transmission must exist (internet, etc).
I think the important thing that tips the scale toward decentralization is the ABILITY to run your own relay, not necessarily that people choose to. I just fundamentally disagree that decentralization means no centralization, at least outside the realm of philosophy or theory. It's just farther from the centralized end of the spectrum. I believe this is why massive regimes always fail as they try to reach peak centralization. On the other end, some things tend to centralize for efficiency. I think Nostr will have the same pendulum like cycles. Obviously, we should try to be more decentralized where we can and when it makes sense.
What would you use to transmit notes that absolutely no one can have any influence over? Because that's what you'd need to have zero centralization as far as I can tell.
Yeah damus is a proof of this. I suspect there will be many other apps and businesses that run relays to support their users and use cases.
I suppose every device running a client could also run a relay, but you still have relays...
without relays there is no async messaging or inbound connections being accepted
the problem is much deeper in the internet than the most important part of nostr's design as an async ad hoc message delivery system
Even in systems theory, some links in the decentralized system can be broken by some node within it. The difference is that not every node is taken out by any single link being broken and everyone has the option of adding nodes and additional links as needed.
I think he's only looking at social media point of view. For that I agree people are used to social media to be free.
But if he wants for example to add nwc for aqua the best way is to have a aqua relay
Instead of using whatsapp for a company group he can make a relay for his company and use something like 0xchat
Most people are seeing nostr just as a social media but theres a lot of the other stuff to explore
Social media is no different actually. Creators making money on social spend enormous money to produce/promote/support their content, adding 10-100$ monthly to have their own relay to make sure their followers get their content is a rounding error. Blue checks are the proof of this.
Median user OTOH isn't going to run public relay/p2p-thing on their phone, that's 100% certain. That's where there is zero incentive to waste battery and bandwidth and space to "support the network" where all you do is scroll through memes for a couple minutes per day, when there are alternative apps without built-in relay that don't produce that waste.
Readers aren't going to pay for the network. Creators will.
I'm guessing the misconception is that all those creators would run public free open-to-all relays and face huge scaling/moderation/legal costs. That's obviously a bad idea and won't happen.
But does an average wordpress site have "comments"? Yes. Does author have to moderate those comments? Yes. All blogs are already tiny "relays" that have some costs that creators pay for. Again, nothing new.
If nostr is adopted, creators will have relays to host their own content, and to host legit reactions to that content - other events just won't be accepted. It will require some moderation and some costs, but that's a completely different thing vs managing a big public relay.
And outbox model will deliver you the twitter-like experience even though the content is scattered across a hundred relays hosted by creators you're interacting with.
Will there be big relays? Yes, twitter will run one, meta another. Will creators rely on them? No, not exclusively. Creators will run their own, that's the whole point.
And if DNS/hosting provider kills their relay, they'll just rebroadcast their stuff _everywhere_, until they set up a new relay in a different place.
Thanks for coming to my TED talk, as they say.
What's decentralization anyway? What's the difference compared to a distributed system?
Isnt decentralization of power just minimizing the degree of trust that is required between system parts? ie not so much about network topology (hub/spoke or p2p). Nostr is rather decentralized in that regard. Its like everyone having their own Facebook in their pocket since every nostr apps verifies accounts themselves (ie authenticity of content). Relays don't have much power. Just a piece of dumb infrastructure like a tcpip router.
That's cleared a couple of things up. So in normies words, instead of paying for a web host for a wordpress site, I pay for a relay. And if it were targeted, as in shut down, I can offload the content across multiple relays til I'm ready to start again off my own back.
If so, why not set up from the get go with different federations of hundreds of relays and via the handy payment system built into Nostr, I am set up to pay like 1 sat/day to each relay runner. That way, in the event of relay takedowns the average Joe doesn't have to broadcast or run around setting up a new relay, and the pros would set themselves back up quicker and easier.
Automatic pay-per-use is nice idea and I want it explored, it just hasn't been built yet, so it's hard to predict how that will factor in. My point is that it's clear who has the incentive to pay for relays - creators, by running a relay or by outsourcing it.
There is an option on hablanews, fountain to split income with selected accounts. That's the basics for sharing your $5 a month paid relay fee across a. bunch of relay runners.
We also do this. There will soon be way more relays, including many self-hosted ones, as use-cases advance. And that's not even accounting for each business having x relay mirrors and/or databases for resilience and geographic arbitrage.
You're missing that he's been away a while and wants to pontificate to the people who never left.
Fair enough. I'm working on doing that less myself.
Disagree with your disagree π
Fair - but encryption aside: nostr is specifically designed to avoid any P2P aspects.
For better or for worse the relay model is what defines nostr π€·ββοΈ - we'll see if that pans out.
"It doesn't rely on any trusted central server, hence it is resilient; it is based on cryptographic keys and signatures, so it is tamperproof; **it does not rely on P2P techniques, and therefore it works.**" - https://github.com/nostr-protocol/nostr
Mind you I'm not convinced that nostr as a concept is actually viable without big centralized relays essentially operating as their own walled gardens in distinct communities
There are incentives for those who run them . Ecash mint operators would run relays. Businesse and creators on nostr would. Specialized Clients looking to provide unique experiences for their users would also run relays
They are not because it is easy to forget nostr means notes and other stuff transmitted through relays. The other stuff seems to be missing in these discussions.
But if the number of users grows, the number of relays will also grow, similar to Bitcoin nodes. The number of relays/nodes increases or decreases depending on the total number of users. I think it's difficult to determine the required percentage.
nostr:nevent1qqs0u0s8y333ywuef0vrmfpr6thrmj78rxjymr3rh95xpffx50qp7tqpz9mhxue69uhkummnw3ezuamfdejj7q3qexcellx58e497gan6fcsdnseujkjm7ym5yp3m4rp0ud4j8ss39jsxpqqqqqqzju853z
Unfortunately I have to agree with this, a relay sits behind a url.
However, you cannot argue Nostr is not decentralized due to relays, take a relay down others will be setup, you cannot stop the sharing of information by taking down relays.
As long as the data that is made is independent of a domain, itβs not centralised and can flow and be accessible anywhere. When data is only accessible from single point, it is centralised. Web5 DWNs can also be considered βcentralisedβ if everyone use a single central cloud provider, but both protocols support any number of relays/DWNs pr user.
There is no chance that Nostr and DWN is going the way of Telegram/Samorai.
Shots fired π
You are 100% on the money. Bitcoin started where every wallet was a node. The Electrum server/client model brought many to the space, but the financial incentives was bitcoin. Nostr cannot succeed without each client as a relay because ultimately there is no financial incentive to operate one. Private and secure bitcoin transaction use was an incentive to run a node. There is nothing for nostr...and it is much more costly. Any device working on this?
nostr:npub1s05p3ha7en49dv8429tkk07nnfa9pcwczkf5x5qrdraqshxdje9sq6eyhe , what do you think about this issue please ?
I agree, however, if the cost of storage goes to almost zero and the cost of energy goes to zero. But the cost of not sharing Pure Human Empathy becomes greater than just using a Fiat Currency!?
The exchange of Pure Human Empathy is paramount, the timing/realisation of that is what makes this so hard to think about.
Bitcoin was given away in an act of Pure Human Empathy, it was exchanged for many years with zero (at the time) exchangeable value. That was a long time ago, but those bands of timelines are exponentially compressing.
The competition is everything for free at no cost (other than your entire liberty at every single imaginable level).
So again, timing is key.
At no time did anyone with any credibility say Bitcoin was an Exchange of Pure Human Empathy. Still not many at all are saying that, but exponentially compressing time lines could be the difference.
Is it the silent majority who actually push a revolution, they are silent because they aren't Psychopaths and are involved in real world exchanges of pure Empathy?
I feel everyone's pain here.