[ 3 / biz / cgl / ck / diy / fa / ic / jp / lit / sci / vr / vt ] [ index / top / reports ] [ become a patron ] [ status ]
2023-11: Warosu is now out of extended maintenance.

/biz/ - Business & Finance

View post   
File: 55 KB, 1020x367, 777777.jpg [View same] [iqdb] [saucenao] [google]
10522173 No.10522173 [Reply] [Original]

just a honest question:
so if they encourage (big) companies to sell their data (data that only they own), then it's nothing more than a centralised network in these cases. Why would someone use a meme network instead a real centralised, if the outcome is the same?

>> No.10522288

Spoiler they wont

>> No.10522291

I thought this was obvious to everyone, all businesses are going to want to have someone that they can hold accountable when dealing with millions of dollars. You didn't really think that IBM would ever want to enter a contract that would be initiated by 15 neckbeards in different basements across the world and where they can just go and twiddle their dicks in the corner if something goes wrong because it's all "decentralized", right?

>> No.10522348

>>10522173
HI

>> No.10522372

>>10522173
>hurrrrrr it's centralized because the data comes from one place

Link isn't about decentralizing data sources, you absolute brainlet.

>> No.10522382

>>10522291
Way to miss the very core of why blockchains allow for smart contracts.

>> No.10522383

>>10522372
Then what is it for?

>> No.10522391

>>10522383
https://link.smartcontract.com/whitepaper

>> No.10522394

>>10522383
Solving the no lambo problem for sergey

>> No.10522396

>>10522391
Use your own words.

>> No.10522411

>>10522396
>spoonfeed me

>> No.10522427

>>10522291
>bitcoin will never take off because people won’t trust “neckbeards in basements” to run nodes
This is you.

>> No.10522451

>>10522396
Think for two seconds, anon.
How in the fuck would Chainlink (or more specifically the independent nodes that make up the network) go about multiplying external sources of data?

>> No.10522470

>>10522411
Ok, it's a thing to allow real world data to enter the blockchain, I get it. That thing already exists (a two line json parser can do this), so what link is actually trying to do is to set a standard for data tampering protection instead, by using some sort of network voting system or some shit, so if Kevin is a malicious node, he gets the guillotine and if John is well behaved, he is rewarded with shitlinks.

Is my homework ok? Because that cannot work for dick.

>> No.10522547

>>10522470
An oracle is in itself a kind of smart contract that implies an agreement on a set of conditions at a certain time to trigger an event.
Multiplying these is not a matter of luxury when it comes to major financial transactions; it's an absolute sine qua non.

https://blog.ethereum.org/2014/07/22/ethereum-and-oracles/
>instead of a long-running contract being run directly on the blockchain, all funds that are intended to go into the contract would instead go into an M-of-N multisig address controlled by a set of specialized entities called "oracles", and the contract code would be simultaneously sent to all of these entities.
>Every time someone wants to send a message to the contract, they would send the message to the oracles.
>The oracles would run the code, and if the code execution leads to a withdrawal from the contract to some particular address then the oracles circulate a transaction sending the funds and sign it.
>The approach is still low-trust, as no single oracle has the ability to unilaterally withdraw the funds

The broader and more diversified the consensus, the more "trustless" the triggers (oracles) are.

>> No.10522566

>>10522451
hurrdurr
the answer on the question why is cl better than [insert random oracle network here] was always that cl is decentralised and there is no competition, they are the golden standard. s-so n-now you say it's o-ok if they are not decentralised?

>> No.10522573

>>10522566
Oh, CL is decentralized alright.
It's just that the type of decentralization you seem to have in mind makes absolute zero sense.

>> No.10522586

>>10522547
How do you reach data synchronization between oracles? Are they connected somehow?
What if a contract demands a minimum amount of oracles to execute itself and there are no enough oracles available?

What about 50% attacks, where half of them say the square root of 2 is y and the other half say it's x?

>> No.10522605

>>10522173
There's no single point of failure and no one can tamper with the data.

>> No.10522606
File: 99 KB, 998x767, 1501993315176.jpg [View same] [iqdb] [saucenao] [google]
10522606

fear the technologically versed neckbeard

>> No.10522613

>>10522291
>businesses won't trust trustless systems
jej

>> No.10522626

>>10522173
a big company would do this to destroy value another big company derives from worthless data, so as to buy them for cheaper

>> No.10522634

>>10522586
>How do you reach data synchronization between oracles? Are they connected somehow?

"As we explain below, their individual responses are aggregated via one of several possible consensus mechanisms into a global response"
https://link.smartcontract.com/whitepaper

>What about 50% attacks
How do you attack something you can't see?

>> No.10522674

>>10522634
>How do you attack something you can't see?
With a roundhouse faggot.

>> No.10522681

>>10522586
>How do you reach data synchronization between oracles? Are they connected somehow?
answer aggregation
>What if a contract demands a minimum amount of oracles to execute itself and there are no enough oracles available?
The amount of required oracles is set by the requester. If there are not enough available for an API he shouldn't set up the contract.
>What about 50% attacks, where half of them say the square root of 2 is y and the other half say it's x?

consensus is not implemented yet, but for your example (exactly 50/50 split between true/false) I'd recommend always using an odd number of oracles.
Generally speaking, Reputation providers would definitly lower the reputation of the oracles sending the wrong answer.

>> No.10522713

>>10522634
I read it quickly. The aggregation contract and the reputation contract will never work, those two alone are a single point of failure that are so complex to solve is not even worth. Worst of all, you'll need hundreds if not thousands of different contracts for each case, what's the point even

>> No.10522721

>>10522713
I hope you've alerted the Chainlink team about the futility of their endeavour. They'd love to hear from you, I imagine you'd save them a lot of time and heartache.

>> No.10522764

>>10522713
>those two alone are a single point of failure that are so complex to solve is not even worth
Says the guy who had to ask how responses would be synchronized, and then literally said he read the white paper "quickly".

"Aggregation" can be as simple as a majority vote, and "reputation" can be as simple as tallying minority outcomes. So "it's too complex" doesn't fly.
Or they can involve any of the multitude of other solutions involving certification, outlier elimination, etc.

"They're a single point of failure" likewise doesn't fly, since the WP is very clear about there not being a single or singular option when it comes to aggregation or reputation contracts.

>> No.10522774

>>10522721
I tried calling them but they were too busy buying them lambos to allure sensitive trap whores.

>> No.10522777

>>10522674
t. Chuck Norris

>> No.10522863

>>10522713
Who is Ari jules.
ex Chief Scientist, RSA,