/biz/ - Business & Finance

>>13926876
wtf I'm selling right now

>>13926876
what? i dont get it

>>13926876
I never thought of that. Thats unironically a good point. Assuming anyone uploaded the data you need within a reasonably recent block.

>>13926947
this works for all blocks because proof generation is with regards to whatever block you want. The state tree contains all state, not just state set in the newest block.

>>13926909
so this whole link thing is about retrieving off blockchain info for use in smartcontracts. But once someone uploads that info, anyone can grab it for free without using Link or paying anything. All reads on ethereum are instant and free, because you are just reading info that is already on your hard drive.

So if someone wanted to, they can just piggyback off other peoples data requests for free without ever sending a request.

So lets say 200 different smartcontracts wanted to know the state of the S&P 500. If any one of them send the request, the other 199 contracts can get the info for free. There is nothing Link can do to stop this, due to the nature of how ethereum works. So even if millions of smartcontracts did start to use link, you still arent going to make millions of dollars. Perhaps a hundred updates a day might get pushed through.

>>13926902
>>13926909
OP is saying that everyone can use the same piece of outside data, with only one entity having the cost/payment...effectively lowering the cost significantly from what it was theoretically and making the race for adoption into mainstream commercial use that much more efficient/necessary.

Aka this will save even more thousands of dollars (at the expense of jews that control (((data)))) like faceburg) then was previously thought...making it that much more necessary to implement for businesses to stay profitable vs competitors.

>>13926989
yeah I know I wrote a smartcontract for ETH once. Im pretty familiar for how it all works.

>>13926909
He means if someone makes an oracle call someone else can steal the answer for free. Its not that much of a big deal the priority was that the informations itself was reliable in the first place.

>>13927009
>>13927025
What does this mean for the link token?

>>13927013
I thought you were referring to the 256 limit of the blockhash opcode, which is easy to circumvent.
Anyway I thought about it and a cheaper option would be to use blockhash + block header + transaction tree hash + the signed oracle transaction itself.
The proof should be much smaller than the state proof because the transaction tree is smaller by orders of magnitude

>>13927072
It means link is not needed even if link ends up being needed.

>>13927082
damn...can any anon debunk this or is this legitimate fud?

>>13927132
They'll just scream fud, brainlet, or no spoonfeeding. Watch.

>>13927132
It’s all bullshit. “All it takes is one on-chain request” yea except the node is querying market data that is changing every second. OP is literally talking out of his ass.

>>13927143
I've been holding link for close to 2 years, but this is the first piece of fud that actually worries me because it actually makes sense.

>>13927132

Legitimate fud, i asked about this a while ago on discord and on telegram, none of the devs responded.

Not sure how much would the token nominal value go for, but its def not gonna make anyone rich.

>>13927072
My best interpretation is that link token usage, and thus demand/price, will initially be more influenced by the number of unique jobs. At this stage we just have eth/usd data being generated. Then maybe we add btc/usd data. Then weather data from major cities. Etc etc until the network grows and we end up with thousands - millions of one-off unique jobs as they pertain to shipping/insurance/derivatives. What’s an interesting idea is wrapping these “base level” calls inside more complex smart contracts so that they have the most recent information possible. You could build a fairly full proof and automated fiat-crypto-fiat payment service with no users aware that crypto is being used.

>>13927155
Isn't the OP right about the answer being available for everyone to see though?

>>13927166
But I’m a brainlet someone else can probably explain it better. Plus price changes every second, not every block. Frequent calls would be needed for accurate pricing

>>13927166
Hmm...so you think that even if the OP is correct, the link token will still be valuable in the long run?

>>13927155
Nothing stops you from grabbing the data from block X and generating a "stealing" transaction with proof instantly so that it ends up in block X+1.

>>13926876

Are you this much of a mongoloid? Dude, we know this already. The point of LINK isn't so people can pay for data. It's so that they can pay for accurate data. They want to know that today TSLA closed at 188.22. They DON'T want to know it closed at 1289172123897LOLJEWSDID911, which could cause problems with smart contracts with garbage in, garbage out.

If the LINK network proves that today's closing price is 188.22, that's all that matters.

>>13927207
Yes but if that accurate data is readily available, wouldn't that really undermine the link token value? Or am I just being stupid?

>>13927165
let me make sure im following. so essentially an intelligent client could check to see if a previous request was made within an allowable time span and then just "steal" the response data. only if no such request/response exists would it then create its own request

>>13927143
Have you worked in any larger capacity than doing your mums laundry?

You think whats high on the agenda for corporations, is saving a penny by using potentially inaccurate or deprecated data on contracts worth many hundreds of thousands, or millions, or more, instead of just requesting their own specific data from specific source at specified time?

>>13927174
cant you just google S & P 500 as well to see it?

>>13927207

Clarification: yes, everyone who can see the LINK network will see that TSLA's closing price was 188.22 for what OP said. The value in LINK is that the network proves that it is the right price and not some pajeetery to cause smart contracts to maliciously trigger.

>>13927229
>check to see if a previous request was made
This takes energy/ cpu cycles

>>13927222
my gut tells me there's an infinite amount of different off-chain data, so theoretically there is no cap on oracle calls, but this would still severely reduce the number of calls across the network regardless. but idk im pretty sure im a brainlet

It looks like it's grabbing the price of ETH from CMC and crypto compare. Don't those sites already aggregate the price from various exchanges via their own calls? What value is CL adding in this scenario?

>>13927188
Checked. Absolutely. While I think this data sharing thing will occur I don’t think it’ll happen as much as speculated. If bargain bin Bobby is working on some project that just needs a rough map of eth pricing, for example, he can look up a delayed response to save some cash on his project. If I’m running a payment service though, I’m going to want to make sure that everything is being converted at the exact right price to avoid losing money. I’d gladly take that few cents fee and just pass it on to the customer as a transaction fee.

>>13927268
yea i dont understand what would go into looking for specific previous calls. this is pretty interesting discussion tho

>>13927222

Because accurate data isn't readily linkable into Smart Contracts. Yes, I can check Yahoo Finance for the price of stocks, or other financial data. Maybe some dumbass pushed the wrong button and suddenly it shows TSLA is $1 million per share. I, as a human, have common sense and think, ok, some dude just fucked up, they'll get it fixed in a minute or two. If you think I'm making stuff up, look up the 2010 flash crash. Mispricings can and do happen. That's an extreme example but the market is just humans and computers who were programmed by humans, so shit happens.

"Smart contracts" aren't smart. They'll just execute no matter how erroneous the data is. You need LINK to sanitize and validate the data first before it is suitable for smart contracts that are self-executing with no recourse.

>>13927286

True, but you aren't the target market. The target marker are major financial institutions who absolutely cannot afford to screw it up.

>>13927229
Yeah but good luck trying to calculate what the outcome of this will be

>>13927240
I see, but if multiple sources were requesting that same data, that would make the link token even more valuable right? As opposed to having one source request the data then having that data readily available for everyone to see.

>>13926876
linktards can now request the price of LINK from a chainlink node and watch it die instead of watching it die directly from coinmarketcap.

>>13927143
>>13927132
Stale data isn't valuable brainlets. If you want to execute smart contacts with the velocity of money today sloshing about between banks and other actors you need data on demand. That is literally the point of oracles. Sure if multiple actors want the same data they can piggyback and chainlink can still facilitate trust during the exchange but only during a specific timeframe. Fewer less time sensitive calls might also make use of old data but why though if it isn't the latest? Like oh someone called for this data last week and it's on chain for free but isn't the latest so we loss money while trying to save it ..

>>13927303
...which means the link token has great value

>>13927286
oracle calls inherently have latency, at a minimum you have to wait a block between making a request and getting the answer anyway.

>>13926909
>>13926876
This meme is reversed, when Toby gets ate by spooder and swoll his glasses make shit blurry you

DUMB FUCKING IDIOT

>>13927295
>You need LINK to sanitize and validate the data first
I thought chainlink was just playing the role of a mailman? Wouldn't verifying the data itself is correct require technology far beyond our current capabilities, like an AI?

>>13927345
Lol no

Read the second blog post

>>13927345
>>13927355
"Verifying" means two different things here.

I believe the first post is assuming validating that the returned value is in fact valid, where the second post is just asserting that the value is true based on it's source.

Or maybe i'm really high biz.

Data that's a week old is worthless, sure, but something that's 5 minutes old? Not saying there isn't a need for accurate instantaneous data but the use case would be significantly more limited.

>>13927345
correct
>>13927355
brainlet

>>13926876
kek. saved

>>13926876
Was just watching the video where Sergey actually talks about this the other day. https://www.youtube.com/watch?v=ytv8U0bejPA&feature=youtu.be
50:40 is where he starts talking about it.

>>13927414
This is what I'm worried about, I have no doubt that link tokens will still be uitlized, but if what the OP is saying turns out to be widespread then that would severely undermine the price of the link token since it's usecase would be nerfed a great deal.

>>13927185
Man you really are a brainlet. The fastest you can possibly upload data is once per block. There is no way to do that faster than once every 20 seconds.

>>13927238
Its not enough that you as a human can see it, it has to be in a format that a smartcontract can read, and in a trustable way so no one can spoof data for their own benefit. Thats the entire point of link.

>>13927268
Smartcontracts cannot "call" themselves. It has to be initiated by an outside computer. I.E. Something not on the blockchain. Look up "ethereum alarm clock" for more info. Meaning that the energy/cpu cycles wouldnt use ethereum as gas. So its basically free.

>>13927450
I don't think so, I think he talks about oracles copying answers from other oracle and pretending they did the work.

>>13927286
>few cents fee
nigger its 2 dollars. And thats at todays prices. If ETH keeps going up it will be 10+ dollars per check.

>>13927316

>>13927710
Re-read what you wrote. Maybe if you do that enough times it will click for you

>>13927786
Non blockchain operations cost trillionths of a penny. No one is going to care about that.

>>13927191
Yes, but the whole point is that the data in block X is from the past. There are certainly applications where your workaround has valid applications but the use cases targeted by CL specifically benefit from the utilization of off chain data, not data which is already available within a block.

What the fuck? This is horrible.
Why did I buy link? Jesus.

>>13927003
holy fucking KEK. how is this kind of scam legal. fucking NEETs get fucking REKT

>>13926876
Fuck you op, why didn't you post this before mainnet? You were sitting on this info, waiting to unleash it at the worst possible time.

Sounds like a huge security issue for any corporation using sensitive data

>>13927003
>using data that's old
>for LINK's use-cases

vringe

Sergey mentioned this very issue in one of his interviews, can't remember which one though

>>13928643
>hardcore Linkies like to FUD

woah...

>>13927295
flash crash didn't happen because "some dude just fucked up" it was the algo trading that fucked the prices up. it was a perfect storm.

here's a tip for all the link bagholders - sell your shit now and buy BTC ETH XRP LTC BCH. not sure why you all chase the next big thing when a mix of the aforementioned coins will get you 100x in 10 years or sooner.

>>13927003
You're right that once the data is committed to the chain, it's effectively public domain. There are a few things I think you're overlooking, though. And I guess I should preface this by saying I'm not heavily invested in LINK -- I'm not quite sure I buy its implementation, but the problem space is absolutely an ideal target for blockchain development.
Perhaps the simplest consideration is about latency of the data. Yes, if someone else had published ETH/USD price data 2000 blocks ago, I can use it, but it doesn't do me any good in terms of knowing the current price. Data needs updating, maybe even at as high a frequency as Ethereum will allow. Another factor is about the reliability of the data -- a major part of Chainlink's value proposition is that it can ensure a degree of trust in the data's accuracy. I'm not really in a position to explain how that's meant to be done (other than, from a high level, Nakamoto-style consensus).
I'd say those two are probably the biggest cases targeted by Chainlink, but there are a couple other scenarios. What if the data is published in an encrypted format, because it's proprietary in nature? (The value of using a public blockchain at all in this case may be dubious, but it's possible nonetheless.) Or what about "niche" data topics -- sure, there's going to be plenty of people that all want to consume the same set of cryptocurrency prices, but what if someone wants to use the Ethereum blockchain to back their marketplace for some obscure WH40K ripoff? (Look up Zipf distributions.) There are definitely still use cases for Chainlink, even if it's not designed to be Bring Your Own Data.

>>13928743
How is it FUD if it's true? Is it wrong?

>>13928667
Link is (presumably) intended only for public data.

But yes you are right if it was private data anyone would be able to read it. So that basically makes link impossible for sensitive contracts. All of ethereum is this way btw, this isnt specifically a link issue. This is why people were trying to push Enigma back in 2017.

>>13928764
>What if the data is published in an encrypted format, because it's proprietary in nature?
that would prevent its use in the smart contract in the first place.

>>13926909
>hur dur link is shit

>>13928764
Its impossible to do encryption on ethereum. There would be no way for the EVM to perform operations on it while encrypted. Such a thing is called Homomorphic encryption and is entirely theoretical.

>>13928796
sgx is core to utilising public chains with private data

>>13928667
pretty much this. no one but retarded 4chan NEETS fell for this scam.

>>13928667
>what are zk proofs

well this is only ethereum. We still have Tron, Cardano, Hashgraph, etc

>>13928816
I thought so too but intel's security turned out to be such a disaster it's no longer a viable proposition.
Maybe if in 5 years after they somehow manage to not have a speculative data leak sgx could be considered again.
>>13928829
this is true for all of them at once

>>13928829
All public blockchains will have the same piggybacking weakness. They might have lower fees though :>)

>>13926876
I would rather purchase from the manufacturer of a product, in most cases, rather than a distributor, and especially a drop-shipper who is just making a quick buck-or-two on the sale. The relationship you establish the manufacturer is key - a trust relationship.

>>13927191
Yea there is. Research more, you aren’t as smart as you think.
The biggest indicator of a trusted node is:
1. UPTIME
2. RESPONSE TIME

If your node is querying data from a previous block, your node will eventually be selected as an outlier for bad response time, as every other node is querying from an API and delivering the data in >5 seconds.
Again, you’re talking out your ass buddy.

>>13928764
so what do you think the implementation problems are then friend?

Also,Have You Really Been Far Even as Decided to Use Even Go Want to do Look More Like?

>>13928876
>t. completely missing the point
it's not about nodes earning link, kek.

>>13928882
shut the fuck up you idiot faggot autist NEET.

>>13928871
are you sure you are in the right thread?

>>13928876
Its not about nodes, its about the contracts themselves going after identical information

>>13927003
B T F O

>>13928828
Agreed but why would I need link oracle if I’m doing zk proofs implementation for off chain data

>>13926876
wtf just sold all my links what a shitcoin

>>13928916
Reeeeee

>>13928926
Yeah I'm saying people that cannibalize the original data, or are resellers or reproducers of the original data, are not as valuable to the customer

Guys I'm autistic please tell it to me straight, I've held for 2 years. Is it really bad or will we recover? I knew I should have tethered but I'm too deluded. Seeing biz panic is really getting to me.

>>13926876
Good meme anon. Except Peter Parker didn't need his glasses in this scene. Link $1000 eoy checkem

>>13927240
The value is in the insurance of staking penalties . Sure you can listen to data coming in, but how will you be compensated if something goes wrong.

This is the most pseudo-intellectual thread /biz/ has had in a while

>>13929149
Buy gold

>>13929141
ok fellow brainlet i will spoonfeed you, this one is easy. give me an example of someone who has the "original data"

>>13929149
X is the number of smartcontracts using off blockchain datasets.
Y is the number of unique datasets.
T is the necessary refresh rate of the datasets.
A is the amount of acceptable delay in the data, in blocks
L is the Link cost of requesting oracle data.

People thought the profits from owning Link would be X * Y * T * L

But since requester can piggyback off each other for data within the same dataset the profits are actually just Y * (T/A) * L

Link may still be valuable, but only if the number of unique datasets is very high. Keep in mind public blockchains cannot be used for private data. So the data must be both unique and public.

It would also help if the data needs to be refreshed a lot. For things like a prediction market, ( I.E. did Trump win the election, for a contract like Augur) one upload is all that is needed forever. For something like the stock market a much higher refresh rate is needed. However for long term options or something long term like that even one refresh a day might be enough.

If this is really bad or not depends almost entirely on Y. It sure as shit isnt good though.

>>13927003
rape. think about it for a second. does it appeal to you as something you will do in your life? now you know yourself.

F flush this coin

>>13929328
This is an absurd assertion when you take into consideration that, as you point out, your example of piggybacking data is only possible for public transactions that don't necessitate real time or unique data inputs. Additionally, anyone scripting contracts that share a single input would be intentionally circumventing the additional degree of security that is offered by the link network's reputation and penalization mechanics. I'm not saying that you're technically wrong, I'm saying that you're retarded if you think this will have a substantially negative impact on the token's future value should its intended use cases be realized.

>>13929618
you are still using link's network so it should still have the same security. And I dont see how you can be penalized for this. This is a contract thing, not a node thing. Contracts dont do any staking or anything as far as I know.

>>13927009
Jesus these mental gymnastics. Companies don't want all their activity public and they don't want to pay for parasitic competitors to leech.

>>13927316
The perfect usecase, a self-reflecting JUST.

>>13928866

JP Morgan Quorum has ZK proofs and that’s what big businesses will use LINK with

>>13929882
>JP Morgan Quorum has ZK proofs and that’s what big businesses will use
[end sentence]

>>13929719
Look man like I said you're not technically incorrect but the examples where your identified exploit would be applicable are limited and implicitly low value in nature. By penalization mechanics I mean that anyone opting to piggyback data would lose the ability specify a penalty deposit which would act as a form of collateral & thus incentivize good node behavior as outlined in the FAQ https://docs.chain.link/docs/faq.. What you're suggesting is ultimately just a way for smart contract developers to avoid paying a network utilization fee, although given the limited cases in which your exploit would be possible I can't imagine that the associated fees would be anything but negligible and therefore have a negligible impact on link's overall value.

>>13929947
Whoever is requesting the data would still be specifying a penalty deposit.

>>13930024
And a high penalty deposit would imply a relatedly high fee paid to the operator for the fulfillment of the data request. Sorry mate I just don't see a case where your hypothesized off-brand contracts would have a negative impact on link's value.

>>13926876
why would anyone buy link tokens and make a smart contract to find out the price of eth when it's clear as day on tradingview
what a useless shitcoin
definitely going to zero

>>13930136
well that was just an example because the makerdao price feed is already available for free.

how low will LINK get?

>>13930473
0.01cent

>>13930473
I honestly think it's worthless, but I think that about a lot of tokens/coins that currently trade high.

>>13930503
wow so funny child

>>13927009
nobody is going to trust a contract thats piggybacking someone elses data. only hobby contracts will be doing that.

>>13930503
>>13930511

Have sex incels