We Should Get on Urbit Holla Forums

with all the accounts on twitter being shoahed, alternatives like gabai and gnusocial of dubious value, I can see some desperation. Theres a couple threads right now where people are talking about social media reform or nationalizing twitter as ways to fight back against this. However using government to get the things you want is obviously a double edged sword as you all should know, so preferably we should be able to get what we need by purely technical means without resorting to such a big hammer.

Introducing Urbit
there's a podcast interview of the urbit creators where they explain it here:
soundcloud.com/arthurfalls/the-ether-review-46-smoking-dmt-with-mencius-moldbug
or you can read their site here: urbit.org/

but to give the you basic gist of what it is and the philosophy and motivation behind it

This shit is truly the future.

Right now theres not a whole lot going on in urbit, there's only one talk channel (basically the urbit IRC equivalent) where the devs chat, but it has literally everything we need. We can start up a federated chan clone without too much trouble.

Come homestead the new internet with me Holla Forums.
urbit.org/docs/using/

Other urls found in this thread:

youtube.com/watch?v=gqGEMQveoqg
github.com/Fang-/Urbit-By-Doing
twitter.com/NSFWRedditImage

Use irc/xmpp

There are more than enough trusted solutions as it is, we should stick to that.

There's more and more a push of new solutions, made by nobodies, with fancy names, logos and websites, that promise all sorts of crazy things that all collapse when scrutinized. It's actually worrying. One could think (((they))) are trying to replace old, proven, trusted, open source software with newer, untested, controlled, often closed source software. Or it could be independents trying to launch the next big thing, and, inevitably, be at the center of it, for money purposes. Whatever it is, the intents are in someone's advantage, which makes those unfair, and something we should stay away from.

The ideal replacement right now is newsgroups, which is easy to deploy, and that we can secure without having to resort to stuff like cloudflare. One of us takes responsibility like moot and hotwheels did before, hosts the thing, tries somehow to have it financed, and we're good.
On that topic, I don't even trust NNTPchan. I'm ready to believe the guy that started it is honest, but it rests on a custom server, and not only it is non standard, it's also unproven. A completely standard solutions allows people to connect in a variety of ways, with different software, each having different security holes, so you can't take it all down at once.

No matter where we end up next, it will be centralized, there's no avoiding that. No amount if computer wizardry will change that. We're still on the internet, we connect through our ISP, contact servers thanks to a DNS, whose IP is attributed by a central authority, and even "decentralized" networks are subject to that, on top of having one, or at best two, central nodes to enter their decentralized network.

Unless you're pulling your own cables, everything is centralized. I've been thinking very hard about the issue, and came up with plenty crazy solutions, and it's even my job as a software architect, but it's still centralized in the end unless we go radioham.

Be careful little user, it's dangerous out there.

radioham is actually a real possibility for urbit in the future. the whole system is architected with the networking integrated with the OS. think, all the internal connections are identical to connections between computers, and this is all overlaid on top of a global encrypted partyline. basically you just release packets that anyone can pick up and whoever has the encryption keys can read it.

there's this guy Van Jacobson who talks about content centric networking here: youtube.com/watch?v=gqGEMQveoqg where instead of designing protocols around the idea of connecting two specific nodes together, the network routes the right packets to the user based on its content, regardless of where it came from. You can see the potential of this idea for radical decentralization, and urbit is a realization of this idea.

just check it out user, if I can get even one autist here into this Ill consider this thread a success.

It's not such a black and white issue. 10% centralized is much better than 100% centralized. If all they can control is your bootstrap point or DNS then they still can't control content, enforce censorship etc. All they can really do is prevent people from connecting in the first place. It's a huge improvement to what we have now. One step at a time.

hmm

I'm more of an IPFS man myself. It's got fairly reliable p2p, provisions for non-internet transport layers, public-key authentication of peers, ability to search for peers with a given object, a http server interface, and not much else. That's all the building blocks you need to make a communication medium that'll last.

On the other hand I struggle to understand what urbit is. From everything I've read it sounds like Distributed INTERCAL.

in the interview I linked, they mention WeChat as comparable to what the end product might look like on the normie end. A cyber-ego, is maybe a good way to think about it, but WeChat is slave-ware, Urbit is for freemen.

Urbit is an interesting idea wrapped in a cloak of autistic nonsense +2. It's true, the concept that Tlon are pushing could truly revolutionise the way we use networks, but they're terrible at explaining it (read: normies don't give a shit) and the purposeful obfuscation of the concept makes it frustrating to comprehend even for other programmers. All the good ideas are veiled in stupid conventions and awkward engineering decisions that seem purposefully designed to be as inaccessible as possible. Go look at the whole crap with the "galaxies" and "consuls" - Urbit is structured like some galactic Roman senate and is frankly cringeworthy. It's a damn shame too, because I respect the concept.

But then, it comes from Mencius Moldbug aka Curtis Yarvin aka the NRx jew who brainwanks to an all too willing audience craving his kosher cummies

pretty sure he did this to keep out leftists and normies from his project

What the fuck is Urbit?

One of the great things about twitter is that you have access to TPTB.

Urbit better be good since Moldbug stopped blogging to pursue it.

OP, I have been going on there sometimes, but I'm too pleb to contribute anything useful to urbit.

Urbit really is the future. Curtis is seriously one of the smartest people alive. He's been working on this for more than 10 years. Definitely gonna see some good things I reckon.

Peter Thiel donated to Tlon btw.

I think for 99.9% of people they won't even slightly understand how urbit works, just as they don't understand how the internet works. But for them it'll act as an aggregator of their social network accounts etc so they'll sign up eventuallh

It is time to go out.
Say what you fucking want, but currently we are on the best part of the pendulum. With Europe almost Right we have to stop taking the streets, normalizing in the head of the normalfags what is really happening and who their enemy really is.
Trigger them online and spread how they are shutting down dialogue.

They expect one of us in the wreckage

by that don't mean killing people but spreading the truth face to face, some might get fined and even sent to prison, but that will only be proof that you did your part in the great meme war

*start taking the streets

Holla Forums here

DO NOT INSTALL ANY SHIT THIS SHILL ADS FOR.
We purged his threads from Holla Forums yesterday. He is also the ToX shill who spammed all boards few months ago in order to get people to use this flawed piece of shit.

Urbit is a bunch of good ideas hidden under a shitheap of self-aggrandizement and sloppy thinking, which perfectly mirrors Yarvin's writings if the parts I have read are anything to go by.

It claims to be a clean slate, but its foundation, Nock, is completely dysfunctional. Nock is unusably inefficient — decrementing a number n can only be done by counting up to n-1, imagine doing actual work with it. Yarvin acknowledges this and proposes (the only) way to fix it: So-called jets, which attempt to recognize formulas and make use of the underlying environment (hardware, Unix, …) to speed up the execution. However, if you ever actually use jets, you don't have a clean slate anymore because you depend on something below Urbit. The most disappointing thing about this is that Yarvin doesn't even seem to realize how jets compromise what he was going for.

If you want to know why I am focusing on the clean slate part so much, you should note that this idea is the reason for all the obscurantist terminology in Urbit. Yes, the reason why Urbit's docs are nearly incomprehensible is an idea that got sacrificed because of its implementation's inadequacies on day one.

Finally, his outfit was bought by some SV venture capital dude, so I don't expect these problems to get better any time soon. A shame, I liked the idea of a system as a recording.

Fuck off back to the cathedral chomsky

Moldbug did nothing wrong

Bump

checked.
What about zeronet

Urbit looks neat. Here's a couple more services:

They're not but that normalfags think so is nothing but a good thing. If you're such a basic bitch programmer that you've never even confronted a revisionist project built from the ground up, they don't want you involved. The meme spewed by all the redditors and hackernews fags of "unnecessarily obfuscated" completely stems from their decision to write a new language not based on old conventions, which is very easily justifiable, and other results of attempting a fresh start. The project is considered hard to explain not because of their documentation or processes but because of the ambition of the project. sA a replacement to the core of nearly all computing as we know it - UNIX - we can hardly imagine what the end results might look like when put in practice and will necessarily result in radical new approaches to areas that are often left untouched.

In any case, there's not much to see here. The most you can do is either contribute to the project or prepurchase 'real estate' for future addresses last I checked.


This criticism is much more legitimate, I don't know how true it is.

I keep wanting to get into Urbit, but everytime I do get around to installing it and creating my planet, the syntax of the languages turns me off.

I'm a long time programmer, nor a stranger to Lisp or Scheme. But the syntax of Hoon is just too eccentric for my tastes. Why couldn't Mencius have just stuck with using traditional s-expressions? Scheme is clean, but every-time I try to use Hoon, I come away feeling dirty. There really needs to be a Scheme-like to Nock translator available for those of us who refuse to learn new tricks.

ToX and Urbit are nothing alike, nor run by the same people. You sound like a shill.

you have awfully strong opinions for someone who has clearly never used it. jets work fine as a clean slate solution because they're standardized. You never have to write your own jets unless you're doing something that's never been done before and you absolutely must have a little extra performance boost. When people start writing video games on urbit, get back to me, but even then, note that people who do heavy optimized work will rewrite their codebase multiple times anyway.

Its a clean slate in the sense of being semantically clean, part of the purpose for this is to put an end package-manager hell forever. And it succeeds remarkably. It's magical being able to see kernel updates happen automatically over the wire with no reboot neccessary, in exactly the same way as lowly apps.

Now you're just straight up lying. Shill or angry nerd? They've sold some galaxies to various venture capital people, that fact that you don't know the difference just confirms you don't know what the fuck you're talking about.


It has everything we need for our purposes though, it works just fine for shitposting. There's chat and forums, the most we would have to do is tweak some css

the type system is actually pretty nice once you get it. a lot of things about this project are strange and frustrating on the surface but actually deeply thought out underneath.

You should check out this tutorial and give it another shot: github.com/Fang-/Urbit-By-Doing
the most frustrating thing in this guide is probably understanding the "moves" which send data between apps, but once you realize its the only idiom you have to use for sending any kind of data anywhere it's kind of astonishing.

I don't need to use a system to see problems with its design. The entire point was that it isn't semantically clean and, worse, can't be clean because Nock is a foundation made of sand. It doesn't matter that I don't have to write jets personally to make Urbit usable for me; the problem is that someone has to do it in order to make Urbit usable for anyone. Before you claim (like Yarvin) that jets don't matter because they are equivalent to Nock: That's playing pretend. Your old x86 isn't basically a Lisp machine because what it's doing is equivalent to lambda calculus.
Foundations matter in particular because they influence the later design. Since Urbit is unusable except if implemented on a different system, the claims of a clean slate are insulting pretense.

If you had read my post carefully, you would have seen that I agree that Urbit has some good ideas. I just object to everything around them.

You appear to be right about the VC stuff, looks like they only pitched to that group. However, the problem is the same: Once you start chasing VC, it's usually curtains for any groundbreaking development, which includes a change of foundation.