URBIT and YOU

Urbit, developed by Curits Yarvin aka Mencius Moldbug, could be the solution to the shilling problem that our poor pollack brothers have. Does /tech have any opinions on Urbit?

urbit.org
youtube.com/watch?v=I94qbWBGsDs

Other urls found in this thread:

youtube.com/watch?v=eEjxNGiXy_c
youtube.com/watch?v=_acTt4_IXYM
twitter.com/NSFWRedditGif

It's intercal for workgroups

After watching 20 min of that presentation, it's basically a Lispfag's wet dream. In other words, mathmatically perfect but practically infeasible.

It's very interesting. If implemented properly, it could be a real game changer. Not going to lie though, a lot of the YouTube video went way over my head. Have a bump because I also want to see what Holla Forums thinks.

I kind of wanted a Proof Of Work as Identity solution to the shilling problem.

e.g. Generate identity = 12 hour work on normal PC.

That way, there's still anonymity but it's easy to filter out the shills (just block the identity they've spent 12 hours generating). Of course, downside is that you have to spend massive resources on generating an identity in the first place. But it'd definitely make shilling managable imho opinion fam.

12 hours might be a bit much though?

youtube.com/watch?v=eEjxNGiXy_c
youtube.com/watch?v=_acTt4_IXYM
looks practically feasable to me. I don't know what you're talking about.

Nah 12 hours is probably about right.
It will keep post qualities pretty high tier.

That idea could work for an imageboard too. Could have, say, bcrypt workfactor 10 for secure tripcodes, publish a randomly generated string, and the first 40 bits have to match to be allowed to post.

...

Oh fuck off. This is about his technology. As far as his politics are concerned, his diagnosis of the problems are pretty well done, even if I don't agree with his solutions.

However, this isn't Holla Forums. If you're triggered by talking about a Jewish tech guy, what are you doing on a board in love with Richard (((Stallman))).

The problem with proof of work systems is that people with lots of resources can also do lots of work. You might spend twelve hours generating an identity on your workstation computer, well, they could have hundreds of computers with which to generate identities, so it actually puts you at a disadvantage because you are very likely to have vastly less resources than your adversaries.

I'm a lot more hopeful for "web of trust" systems, but these are very tricky to get right and are necessarily not anonymous.

...

Your fucking retarded.

I really want urbit to supercede other OS's. The technology behind it certainly does, but the "normie factor" isn't there. All they really need is better marketing.

I have no idea what Urbit is, but I've come into this thread on at least 10 separate occasions just for the girl

she's a model or some shit, you daft cunt. Reverse image search should get you her instagram

...

...

Note: I've only read the first fifteen pages (five chapters) of the whitepaper.

Nock is a Turing tarpit and probably the biggest weak point of the system so far. Yarvin tries to get around the obvious disadvantages of such a foundation with "jets", but these are diametrically opposed to the goal of a clean slate with Nock as its foundation. An operation that is implemented outside of the actual system is a primitive. Whether it is equivalent to some piece of Nock doesn't matter; any pure code is equivalent to some piece of Nock. The argument concerns engineering, and for these purposes, Nock is not the only foundation.

This is especially a problem because jets are not limited to simple things like decrements. The whitepaper proudly states that "Urbit crypto is jetted by C reference implementations from well-known libraries.". Said library happens to be OpenSSL. If OpenSSL has issues again, (it will probably do next thursday) how are you going to fix it? You'll do it by going below the lowest level of the system. This is not a clean slate. Sure, you could say the same thing about hardware, but the difference between software and hardware is of a completely different nature than the difference between Urbit and Unix.

This makes the entire thing feel like a game of pretend. It sacrifices the one thing it claims to do (and that thing the world would urgently need) in the name of "practicability". That is not wrong per se — the foundation chosen at the beginning is just bad and forces you to do so.

Its really cool so far. After it got posted on hacker news it had lots of people using it to the point urbit.org went down.
Curtis is very smart, but also a very strange dude...done his fair share of acid I think.
If it ever gets up off the ground properly I can see some really epic things happening.

That isn't true, as a lispfag, I'm not that excited for urbit to replace standard computing because it lacks some of the things that made lisp machines great. The ability to inspect, interact with and step through all the code that the computer is currently running is, in my opinion, an essential element in ensuring freedom for future computer users. No matter how reproducible something is, it's ultimately pointless if the user can't learn how and why it works on their own machine.

The other major factor is having a higher base-level abstraction, that is to say, the furthest thing down that the user has to be aware of, ideally this could be Lisp instructions implemented in hardware as in the Lisp machines. Urbit is particularly bad in this regard, being a whole new esoteric architecture implemented on top of a decaying Unix stack which is implemented on arcane processor architectures.

Even as a fan of his writings as Moldbug, I'm averse to the way he's chosen to present Urbit; its technical documentation is written in the same rambling, alien prose that defines Unqualified Reservations.
Whilst this is worthwhile for a social commentary where you wish to detach yourself from words with existing connotations, inventing new technical phrases for concepts that already exist just makes your work impenetrable for anyone unfamiliar with your prior writing or without a lot of time on their hands. Then there's the fact that Urbit's concepts are rooted in some reasonably high-level CS academia to begin with.

I hope it succeeds and I'll continue to follow it closely, but I'm not ready to proclaim Curtis king just yet.

this guy is very punchable, not just because he's a deliberate asshole, but because he's grandiose at the expense of clarity. I have several questions, I hope somebody takes them at face value and answers them instead of walls of butthurt.

do we really need another slow shitty interpreted language?

if his problem with the personal server is putting a unix box on the internet, why does his software specifically require a unix box on the internet?

how is this different to installing ubuntu into virtualbox compiled as java bytecode running in chrome, running in virtual box, compiled as java bytecode running in firefox on ubuntu level retard abstraction? what improvement is there over a shiny web interface for a regular unix machine like OwnCloud for example?

if the goal is to use the whole network as the storage space, what happens when I turn my computer off for a couple of months and go on holiday? what happens if several important people go on holiday?

what measures are in place to stop people with little interest in the citizen web from hoarding thousands of urbit addresses against the posibility their "property value" actually becoming nonzero?

like bitcoin?

You really missed the point of what he is saying.

You stop putting data in Facebook. You put data on your hard drive. Facebook asks to peek, you're in control.

on what planet does facebook not just cache your data for speed and nefarious ends?

If anyone wants a planet to experiment with no, I have tons. Leave an email itt; disposable is ok.

I'd love one, thanks, user. How did you get a hold of so many?

they're simple to get once you mess around in the terminal enough.

...

Are you doubting that Stallman is a jew? He really is, ethnically. It's not a secret.