Any computer scientists want to explain what's this is saying? I literally have no idea

Any computer scientists want to explain what's this is saying? I literally have no idea.

Is it bullshit nonsense or actually something profound?

It's the white paper of some altcoin shilled on Veeky Forums

Other urls found in this thread:

link.smartcontract.com/whitepaper
twitter.com/SFWRedditImages

...

lmao what coin is this from ?

teaching code monkeys math was a mistake

Pretty sure it's bullshit, but he's trying to say that the joint probability of 3 events is (or should be) bounded by negl(lambda) (whatever that is)

is this from boneh's cryptography course on coursera?

cryptography is hard as fuck. there are like 30 people in the world who understand this shit and 28 of them work at the nsa.

To sparse information to really know what this is.
Looks to me like some sort of proof based crypto, maybe some sort of zero-knowledge protocol stuff?

"Security parameter" usually is the length of the input.

negl(lambda) probably means negligible on the input length, what usually means that the probability decreases faster to zero (in the input length) than any polynomial.

? Are you trolling? Crypto is difficult, especially when implementing stuff in practice. But come on, 30 people? Ridiculous.

You should try reading the entire paragraph instead of just looking at the pictures

chainlink

white paper is here


if anybody wants to look over it who has a knowledge of cryptography ill toss you a bit of ETH or somehting...
link.smartcontract.com/whitepaper

I'm no expert but just seems like a cryptographic proof of security. They are normally presented like that to show that the probability of an adversary breaking the protocol is negligible in some variable i.e. not feasible irl

Don't know what exactly you are looking for and if the following is helpful.

The paper is about building a decentralized oracle network to solve problems concerning communication of smart contracts with the outside world (outside the blockchain). This draws some security issues (authenticity, integrity, confidentiality). This one is a formalisation of the requirements for authenticity.

Read the definition on the previous page (p.24) it is written there whtat it means. Basically: The oracle is authentic, if the probability that an adversary can convince a verifier for a wrong message is negligible.

Some of the non-selfexplaining symbols:

[math] pk_i [/math] : public keys
[math] \sigma_{att}[/math] : Actual signature (don't know what 'att' means, from context probably attached)
[math] \mathcal{A}^{\mathcal{F}_{}sgx}(1^{\lambda}) [/math] : Poly-time adversary with access to the trusted hardware functionality SGX
[math] \Sigma [/math]: Is a signature-scheme with [math] \textbf{Sign}[/math] and [math] \textbf{Verify}[/math] functions
[math] \text{prog}_\text{encl}.\text{Resume(id,params)} [/math]: is a [math] \textbf{url}[/math] in their model.

>Is it bullshit nonsense or actually something profound?
It's not bullshit. Tbh, I've not read the whole paper yet, just the first three sections (but plan to, smart contracts are a very interesting topic and what I've read so far was also interesting). But I got the impression (of course I might be wrong on that) that they just wrote this down to have a "cool" math formula in their paper. So I would say not needed but also not bullshit nonsense.

the other 2 are russian

ew, what an ugly scramble of fonts
tex was a mistake, high quality typesetting does not belong to the hands of everyone

good christ this, if this is what passes as acceptable notation in cryptography, it explains why the field is so unapproachable

>there are like 30 people in the world who understand this shit and 28 of them work at the nsa.
Youre beyond retarded

Good Christ says says encryption on a computer is silly because the quantum computer users in the future can use their temporal modems to decrypt data in the past and then also upload the data in decrypted form to other classical computers in the past
>encryption
>write it on a piece of paper instead of your computer

Actually they are meeting on standards early this year to come up with encryption quantum computers will struggle with. Doesn't matter they still don't exist

It does matter because they already exist in the future and the reality of the universe is based not spacetime, not just space.

If you don't want hackers to steal it, write it on parchment and use the Pony Express

No. Have a cyphered smoke signal essay.

>still being a rationalist in 2018

It's easy when you don't watch a bunch of weird videos that fuck up your mind and make you live in a state of perpetual terror that you lie about being in.

a bunch of autists who think they know computing because they successfully installed Arch. they won't help you with this

With crypto you're not just worried about if someone can break it today, but also if it's still going to hold up in ten years, or twenty, or a hundred.