#nostr is not #Bitcoin

there is nothing original in the design and specification yet, that isn't served best already by cloud architecture.

the resource requirements for relays will grow beyond reasonable home server and small VPS operator capacity. without a mechanism for funding relay operators, and sharding the data they handle is designed, it's on a highway to centralised hell.

even the minimal feature to allow users to specify an expiry date isn't there. the format is JSON which is a very bulky data encoding. it uses HTTP websockets, which are very heavy in the negotiation phase and thus not well suited to gossip heavy traffic, required to make relays peers not just servers. all associated media in notes are currently served already on a very small number of cloud based CDNs.

none of the people involved in making the specification or allowing proposals in have shown any interest or understanding of any of these requirements.

nostr has a chance, over the next year, to get serious about centralisation resistance and after that it's just gonna be another buzzword used to sell the same old shit.

Reply to this note

Please Login to reply.

Discussion

Yes, who says otherwise. #nostr is a protocol that's leveraging on the existing #bitcoin protocol. That's it.

i just literally read an influencoor saying "nostr fixes this". no, not yet it doesn't.

also, it hardly leverages anything. centralised lightning services. lightning isn't really quite ready either without the last mile problem solved for async payments (which nostr needs).

i can assure you, that the hype train has been successful and shitcoinery is about tho descend hard on nostr. i'm probably going to be paid to build some of such systems in the future.

nostrich tears are going to be very tasty beverage.

not saying i don't want to help fix the problem but it's completely unrealistic to say any of the things people are saying. there just isn't many users yet, once the users are here, all bets are off.

Well said

The relays need a forced sync between all other relays who signal they're online. Without this, it could lead to "big relay" scenario of centralised-ish network.

Of course, forcing relays to mirror one another will increase data size exorbitantly as well as prevent niche topic, micro relays from existing. That would likely mean it's not cost-effective for most to operate a relay. But, that knock-on effect is a discussion for another day, once the first bit is addressed.

the "other day" is between now and when amazon, facebook and google take over nostr branding and it becomes meaningless.

it requires the design of a consensus algorithm, and i don't mean "adding a shitcoin and putting it on a blockchain". that one has already been done. go visit steemit.com to see that in action.

i mean something along similar lines as bittorrent, something that is always usable at any scale from dialup all the way to the whole AWS network.

I think you are missing an important point. The decentralization is more importantly happening on the profile side rather than the relay side.

Yes we might end up with super relays. And yes that might seem centralized.

But your profile and social graph can still hop around unlike traditional social media.

You still have the choice to avoid the super relays. You can start your own relay.

Maybe we will have “the relay is in the phone” in a year or two. Which I believe Victor kinda talked about a while back.

it's not that big an innovation to be able to easily migrate your data from one silo to another. facebook and google have been letting you download "your data" for years. i'm sure that a simple application could be made to do a lot of it. sure, the silos don't let you add data to the profile with your chosen datestamps but aside from that it's still the same.

"relay in the phone" means a centralisation resistant protocol that can operate at that level where a tiny node can participate and not be excluded or unreachable.

it really doesn't take that much study of distributed systems theory to understand why nostr is not an innovation yet, celebrating it as a cure for centralisation is extremely premature.

Downloading your data is not like profile interoperability.

My Facebook account means nothing to Twitter. My Google account means nothing to MS Office. My Pinterest account means nothing to Yelp.

Being able to use one key to access all of those is a really novel idea.

I don’t believe this existed in the past.

it's one small step. it's not a centralisation resistant protocol, it's choose your own master protocol.