So Chainlink has been debunked to be a scam, now what?

So Chainlink has been debunked to be a scam, now what?

Attached: Screen Shot 2019-07-12 at 12.47.16 PM.png (908x600, 186K)

Other urls found in this thread:

youtube.com/watch?time_continue=4371&v=ytv8U0bejPA
youtu.be/g8OYsHaOoF8?t=654
twitter.com/SFWRedditImages

Can we all agree on a price point where all you fudding retards just shut the fuck up forever?
Even if it's $100 or something, can we come to some kind of agreement?

cringe

bitconnect reached $300 and still got fudded.

It's time for you to finally have sex

I'll shut up when Chainlink is actually decentralized or has an amount of oracle users in production that justifies $3bil marketcap

when the price is $1000 eoy, that solves the sybil problem

Attached: Screenshot_20190701-194213_Chrome.jpg (810x6138, 2.64M)

how about 0.04$, thats how much it should be worth approximately

I'm kinda autistic about the sybil thing too. And the answer is: for centralized data there's no answer to the sybil attack problem. Chainlink will quite unironically be a bandaid solution but it will be that way for decades. The good thing is: there's real world reputation and the contract owner can choose any node he wants. He could use 3 very reputable nodes and then 100 low reputation nodes as a fall back or something.

How about at $100 if that comes first?

that is non technical bullshit to fool non-programmer normies like youself. There's not a single technical truth/explanation in it

>nodes can be faked
You choose the boxes to call upon in the smart contract nigger I read the documentation.
>KYC is centralization
Except it's not, it places legal liability on the node operators to deliver accurate data. Also, in the future, multiple companies will perform KYC, like Oracle.

>The good thing is: there's real world reputation and the contract owner can choose any node he wants.
You've just described centralized oracles. Something that already exists. Something that does not require the LINK token.

kek you literally have sub 80 IQ. Kill yourself.

1. There's no way to know any of those boxes are legit or not the same scammer, unless Chainlink KYCs them.

2. Legal liability is centralized. Multiple KYC providers is centralized. All these things are not censorship resistant and do not need the LINK token, because they are centralized.

This is already implied in the OP, but you're too low IQ to get that.

req 2.0
sub 250m market cap by 2020 :)

When I dump my bags on Reddit, the Sybil problem will be solved.
In the mean time, I'll give Google, MSFT, and Oracle some time to come up with a long term solution that keeps me holding.

Okay nigger debunk this. Secure enclaves allows calling data from anonymous node operators securely without the data being processed being exposed to them, making Sybil attacks impossible. Also I meant nodes not boxes but I'm not home and phoneposting rn.

Attached: Downloaded the image to Messenger_Lite.png (500x496, 119K)

>entwork

>justifies the 3b mcap

Oh it all makes sense you're just new to crypto

What's the point of staking then? Sounds like a race to the bottom to me in that scenario.

Imagine still fudding after Oracle. Jesus Christ...

>Secure enclaves allows calling data from anonymous node operators securely without the data being processed being exposed to them, making Sybil attacks impossible
They certainly do not. That's just another lie Chainlink says.

>It-it's just a lie!
Nice refutation shitskin.
Even when using secure enclaves, innacurate data can be returned upon request due to technical errors. Staking is insurance against this as well.

There's literally no proof anywhere that TEEs can prevent sybil attacks. You're making the extraordinary claim, you're the one with the burden of proof.

I doubt you understand TEEs, you're just believing the lies Chainlink is feeding you without question.

As for stake, I couldn't care less about a system that Chainlink has not even started designing, thinking about or researching 2 years in. Like I already said, those systems are complex and require years of research and real projects like ETH and Cardano publish papers about it. Scam projects like LINK do not.

Describe how the attestation process works from a TEE in your own words. Answer that and I will have a proper discussion with you,

>no proof TEEs can prevent Sybil attacks
What is Town Crier?
>ADA
lol

Give me a single reason why TEEs are sybil resistant. Hint, you can't, because they are not. Even the Chainlink devs themselves won't defend this point.

Lmao, and your low IQ OP is supposed to be technical? Both my post and yours were philosophical in nature, and yours was babby-tier.

Now you've spent a lot of time in this thread, so i think you probably need to dilate. If you don't, your gaping wound will close and you'll no longer have even a grotesque imitation of genetalia.

stale FUD

Go dilate. “Faking” LINK nodes. Kek.

Attached: 15402FF7-C64E-4661-B398-77A49013BAEC.jpg (530x685, 308K)

yep, never selling

cope, I've explained in detail and proven beyond a doubt why Chainlink is centralized and not sybil resistant.

TEEs do nothing to prevent sybil attacks. If you disagree, prove it, say why. You can't because THEY DON'T.

BTFO

>MUH Sybil attacks
You should be more concerned about SEC attacks.

Attached: 1529162692435.png (3024x4032, 3.95M)

>4channel debunking anything

>TEEs don't insure against sybil attacks
They do, because the data the node is providing is secret even from the node itself. How, then, could an adversary game the contract when they don't even know what data they're providing?

sybil attack is gaming the system by pretending to be multiple attackers, when you are a single one.

TEEs do nothing to prevent that. TEEs can easily be gamed, if they could not, there would be no need for decentralized oracles.

$0.10

Now go fuck yourself.

Or Sergay dumping his 650million StainStinks.

That will be glorious, the day that all Stinkies hang themselves.

>Describe how the attestation process works from a TEE in your own words. Answer that and I will have a proper discussion with you.
From a programmatic standpoint, not a conceptual one.

1. Why would you have that many 0 link nodes? They won't be included in high value contracts at all
2. KYC is optional though, the way the token is planned to work will cover this
3. Collateral and proof of stake does not solve the problem (lol except for creating a gigantic financial hurdle for anyone wanting to even attempt it) because... it's too complex apparently? For who? Oh, this is the part where the writer of the FUD just got lazy and inserted personal incredulity and baseless pseudo-intellectual smokescreening. If they were smart about FUDing they would have at least made up some code or they would have just copy pasted some random code from the github and pretended that it was a big deal to fool brainlets.

THey really don't produce this shit like they used to. I missed old fud. At least some of it was funny, this is just thinly veiled begging.

chainlink cannot even verify tee results onchain. they have "trust me bro" mechanism in place currenlty

Cringed

I'm a developer and I can confirm that op is a faggot

you are faggot and i confirm that you suck dicks

this is what 2 years of built up cope looks like

You fudders aka shills 2 weeks ago are pathetic. No one here is buying your shtick and we all know none of you have sold a single link

Cope

he has to do it very very slowly, a shit illiquid market can't adsorb his exit scamming that ez

Jewish fud

Wasting your time OP these guys are in too deep. I've been posting this clip for a year but it gets ignored because it can't be debunked. Sergey thought LINK was a shit idea but then decided the ico frenzy was too much of an opportunity to pass up on. Watch from 1:12:50

youtube.com/watch?time_continue=4371&v=ytv8U0bejPA

This. Gas the kikes NOW.

HEIL SERGEY

The contract writers will maintain their own nodes!

Bank that want to use smart contracts they will use their own nodes for many functions so they don’t have to pay fees to some other node operators.
Fully trustless is years away, but the banks can still use the technology to decrease costs in a huge way while also increasing trust somewhat. A bank can have its node tied into a number of oracles and be near trustless.

Chainlink can be decentralized or centralized at the same time.

The mega corps aren’t going to trust something like chainlink initially, but they can trust their own node that they operate. Then once the corps have their own nodes in place, they’ve also built supporting oracles. It’s after this that you can achieve trustless state, when enough trusted oracles exist. This is 5+ years away.

I remember that thread where user was phone posting from prison. Good times

>what is Town Crier
You're just recycling old RLC fud now lmao

No one is selling and neither are you. So shut the fuck up with your idiotic fud

someone pls bomb this pajeet callcentre, tired of this discord tranny spam mainnet is what we agreed upon

Never gonna stop :)

TownCrier is not on chain. BTFO stinkies.

Now we wait tomorrow for this very thread to reappear

>Sybil on LINK oracles
Isn't this inherently addressed by staking, the smart contract holder choosing which oracles they want to use themselves, and the blockchain structure itself? Oracles that are all ran by the same guy providing faulty data might fuck up a few contracts at first, but he'd almost immediately be ranked as a bad oracle once people stop staking on his nodes. Then contract holders would simply not choose those oracles in the future based on the bad reputation and history. The final and most important thing is that someone trying to coordinate a Sybil attack would likely be using the same nodes and addresses over and over again, unless they are starting from scratch everytime (which is a bad idea because they'd effectively be running nodes with no reputation). The blockchain itself would allow you to see a history of several nodes that frequently bid together and likely diverge from the randomness that one might expect among different bidders. Doing some statistical correlation tests among oracle candidates could let you sniff out a potential Sybil attack before it even happens, even with candidates that might do things right (to build good reputation) before coordinating an attack later. Just monitor for capital flow after job completion, co-bidding habits,etc.

Link is actually a scam

I will add that XMR could throw a monkey wrench into attempts to monitor past capital flows for oracles, but a SC holder could probably just straight up refuse to use any Oracle with capital flows into XMR or similar coins to stay safe.

except it is lmao

Dilate

retard

>So Chainlink has been debunked to be a scam, now what?

it pumps to 1000, because scams pump hardest.

Dilate

you need to read the whitepaper and have sex first before your incel mind could even begin to understand LINK

>TownCrier is not on chain. BTFO stinkies.

The Town Crier (TC) system addresses this problem by using trusted hardware , namely the Intel SGX instruction set, a new capability in certain Intel CPUs. TC obtains data from target websites specified in queries from application contracts. TC uses SGX to achieve what we call its authenticity property. Assuming that you trust SGX, data delivered by TC from a website to an application contract is guaranteed to be free from tampering. This authenticity property means that to trust TC data, you only need to trust Intel's implementation of SGX and the target website. You don't need to trust the operators of TC or anyone else. Even the operators of the TC server cannot tamper with its operation or, for that matter, see the data it's processing.

Thanks to its use of SGX and various innovations in its end-to-end design, Town Crier offers several properties that other oracles cannot achieve:

Authenticity guarantee: There's no need to trust any particular service provider(s) in order to trust Town Crier data. (You need only believe that SGX is properly implemented.)
Succinct replies: Town Crier can prune target website replies in a trustworthy way to provide short responses to queries. It does not need to relay verbose website responses.
Such succintness is important in Ethereum, for instance, where message length determines transaction costs.
Confidential queries: Town Crier can handle secret query data in a trustworthy way. This feature makes TC far more powerful and flexible than conventional oracles.

youtu.be/g8OYsHaOoF8?t=654

Attached: Annotation 2019-07-03 195613.jpg (2327x1260, 223K)

wait till retards come out with their pajeet responses. but the fact remains that calculations done in sgx cannot be verified onchain by chainlink

also
>trusting intel

it's another way to to ensure security besides just decentralization
if you understood anything about semiconductor fabrication and technology you would realize this is how you know the government is involved
>trusting intel
all processors have backdoors at the hardware level
just like how xerox machines scanned everything that you printed and especially what was shredded hahahaha

Attached: diagram-intel-software-guard-extensions-reduced-attach.png (610x362, 112K)

Attached: Intel+SGX+101_+Isolated+Execution.jpg (1024x768, 95K)

>trusted execution environments

Attached: Trusted+Execution+Environments.jpg (1024x768, 94K)

Can someone get Donald trump to tweet this FUD?

except bitconnect was an obvious scam to everyone except norman

1/2

Attached: BitcoinCope.jpg (2951x9999, 3.2M)

2/2

Attached: LinkCope.png (1350x2700, 3.86M)

mfw

Attached: really rich.gif (480x270, 461K)

Once again, everyone in this thread is a fucking idiot.

The reason Bitcoin doesn't get Sybil attacked is due to the extremely high cost of mining. It is unfeasible to take over the network with fake identities simply because it costs a lot of money. This is also the reason why ETH can not be Sybil attacked under POS: The cost of acquiring a majority stake in the network is too high.

Sybil attacks will be unfeasible if and only if the collateral at stake in the Chainlink network will make acquisition of a majority stake (or at least a substantial amount of the token supply) unviable. This is addressed in the third point of the picture in the OP and discarded as unrealistic because it is a really hard problem, which is wrong. It is a very simple thing.

scam

Let it go, fagwad. This isn’t reddit.

>centralized oracles already exist, and are valueless.
>pic related and op is a faggot

Attached: oraclerev.png (272x948, 126K)

>unless Chainlink KYCs them
Which is what they are working on, they are fleshing out the node rep. system as we speak you low IQ, low energy, faggot following, pathetic pleb.

Attached: 1561391608576.jpg (429x589, 78K)

>debunked as a scam
Nigger loving pajeet confirmed.

Attached: 1561766248194.gif (476x348, 2.94M)

the dilators will keep flooding in
we'll keep blowing them out

Attached: 1561169842432.jpg (1882x2400, 1.78M)

>So Chainlink has been debunked to be a scam
>req 2.0
>sub 250m market cap by 2020 :)

Attached: 1562043688282.jpg (720x583, 28K)

QUICK SOMEONE TELL GOOGLE THEY'RE ABOUT TO MAKE A HUGE MISTAKE

Checked.
My fren 42, hello again.
Imagine not buying LINK for picrelated

Attached: 1562872118055.jpg (1070x2213, 816K)

DID U SAY STEAK?

Attached: huh.gif (257x377, 372K)

It's not LINK invention, it's Intel SGX.

16.5 million people in the world with a NET worth of 1 million or more
when you account for everyone's debts and mortgages that figure makes sense
have a deadline submission today for a client so I'll be mia until tonight
cheers
>mfw I keep feeding off the haters and dilators hahaha

Attached: power up.webm (1280x720, 2.83M)

Godspeed.

Attached: 1561583587068.jpg (206x245, 11K)

mfw those beautiful digits

Attached: 1561094830074.png (503x689, 273K)

>mfw

Attached: 1561566374773.jpg (820x983, 144K)