I literally know of exactly one person excited by the Pear stack, nostr:npub1h8nk2346qezka5cpm8jjh3yl5j88pf4ly2ptu7s6uu55wcfqy0wq36rpev
Quite interesting to never see anything about it in my nostr feed
Speaking of, is Keet open source yet?
Discussion
I listened to the Guy interview yesterday, too and had the same thought. They did not go into much detail. Just that it magically works and that sounds intriguing but I also think I remember that some part of it was closed source and thus a no-go for me.
I want to understand the stack and the promises of the Pear stack. Can I run a Pear enabled nostr relay using a pubkey at home such that I can connect to it with my Pear enabled mobile client from wherever? I bet the devs would claim so and that would be cool but with not much buzz on nostr I feel like there must be something wrong with it.
Yes and the fact that they seem to be an isolated group of devs under one company financing it
paolo who sort out mafintosh to build the team to do this is interested in it because he believes correctly it is the right technology especially over nostr
mafintosh has been at this for a long time before he met paolo, this was not their first project, mafintosh has a long history in open source, pear is completely open and makes up most of keet
they dont want to open all of keet yet because they want to keep everyone updated while they iterate fast, which they have done a lot of, they are also concerned with forks and exploits interfering with dev and interest in keet
there is a lot of interest in pear and there are many projects being built with it, they are usually discussed on keet because keet is now a fully functioning fully p2p ie no relays or servers social platform, it scales infinitely including for media or any files
nostr btw is still using centralized file storage, keet is completely p2p
fun bit of trivia mafintosh created webtorrent
the buzz is on keet, there is no real need to discuss keet on other platforms because keet functions as its own social platform, and unlike nostr there is no relays to pay or configure, and all the files and media are distributed and censorship proof, keet is everything nostr is missing and needs
Yes you can do those things. And everything about the Pear stack is open source. Keet is just one chat app built on it, and it’ll be FOSS when they exit beta.
If “not enough buzz” was a reason to suspect there is “something wrong with it,” you’d have completely missed the first 8 years of #Bitcoin, just sayin. Not a reliable heuristic, imo.
Nice point Guy. You make me wonder if hanging around for long enough around Bitcoin turns people into philosophers!
You either die before hyperbitcoinization, or you live long enough to become a philosopher.
I'm definitely interested, but need keet.io to go open source with an apk on github so I can use on #Graphene
most of the code is open with the rest to be after beta, an apk will come soon
The more I look into the pear stack the more skeptical I get. Most of their hyper packages don't use common dependencies and instead they have their own implementation of low level things like streams
They might have a good reason for it. But it always makes me skeptical when I see other devs reimplementing core concepts
Agree. I think the goal is noble and devs great but their lack of visible traction, the reimplementation of everything and them being so under the umbrella of one company makes me super skeptical too
no one else has managed to achieve what they have which is a social p2p app that scales big, so if they have done something themselves its because they needed to
Not sure why they went with implementing their own versions of things, but Maf and a few others have been at it for a decade. It’s also essentially an extension of his previous projects like webtorrent. He’s just been designing P2P stuff around the DHT model for a very long time. I don’t quite understand the dedication to the append only logs, but for whatever reason, it works, so I can’t knock it that much.
I agree although, I don't understand or agree with their development methods but it works very well for a P2P application and so I give it the benefit of a doubt
I've been using HyperDHT in a project for a while and its been great. its the first library/app I've found that JUST lets you do p2p connections. all the other ones I've looked into come bundled with a bunch of extra unnecessary stuff
I think pablo uses it as well
I’ve gotten rumor of another project that is taking a simpler approach like this and actually removing a lot of the higher layers of the stack to break it down to an MVP like that. No details to speak of but I’m staying in touch about it. Sounds exciting. The second it is public we will absolutely be covering it on 🍐The Pear Report🫡
That’s because all the people excited about it are literally hanging out on Keet and pear apps, lol. I mean if you go to Twitter you won’t find many people excited about Nostr… cuz they’re all on Nostr 😆
Fair 😄, it's interesting nonetheless to see near zero talk about it here. Is nostr brought up in Keet at all? know of anyone interested in building a bridge? I'd love to follow some pubkeys from over there
Also trying to listen to most Pear Report episodes.
Yes actually. Talked to a few people interested in making bridges. Also have talked to some devs myself about building a Nostr client that runs entirely over Pear but we have some other projects to finish first. If we had a separate client that worked really well i imagine bridges would pop up naturally. Of course there’s no reason the functionality for connecting to both relays and directly to peers couldn’t be in the same client.
keet is quite functional enough to carry all its community chat about it and pear on keet
there is a link to the community room on the profile page in keet
today nostr was discussed a lot on keet, supersu the holesail dev is trying out running a nostr relay via holesail
can nostr and pear pubkey formats interoperate somehow?
Not sure. I’ve been looking into finding a path for this with BIP85
i dont remember what the new keet multidevice id system uses but hypercore is using ED25519 https://github.com/holepunchto/hypercore-crypto and https://github.com/sodium-friends/sodium-universal and nostr is using Schnorr signatures standard for the curve secp256k1 which sodium wont do, they are both 64bytes i think