/biz/ - Business & Finance

>>15716850
Public data source gets hacked, spoofed, or goes down?

100 nodes querying 5 different data providers and then aggregating the results transmitted cryptographically secure over a uninteruptable immutable p2p blockchain network?

Ask synthetix what "lel just use a single public api" does the first time it glitches or inputs unsanitized data. (They lost 37000000 eth ).

This isn't for your pokemon knockoff rpg ddap...this is for people not trying to lose 6+ billion dollars to a simple arb bot.

>>15716963
>sole public data source gives out wrong data
>nodes therefore give out same incorrect data

Name a use case where a contract is not executed based on public data (source data may be corrupt) or a single private node from a company (no decentralization possible).

>>15716850
Repeat after me:
t r u s t l e s s

Now go and think about it.

>>15717148
He literally said
>100 nodes querying 5 different data providers

>>15716963
>>15717233
None of this makes any sense. The vast majority of data can and does only originate from 1 source. What’s the point of decentralising something centralized? Why would it benefit me to have 80,000 Chainlink Nodes verifying the data from the same 1 weather recording station in a city? Why would I not just directly connect to the starion’s outgoing feed and cut out the middleman?

>>15717284
lol

>>15717284
Your smartcontract will have to do all the parsing and error checking for the source. What if they suddenly switch APIs? You'd have to make another smart contract. Link outsources that.

>>15717368
>Your smartcontract will have to do all the parsing and error checking for the source

Do you know what a JSON Parser is?

>What if they suddenly switch APIs? You'd have to make another smart contract.

What the fuck did I just read? Are you just making up babble? Firstly, APIs don’t really ‘change’, if they do they’re not sudden. Secondly this has nothing to do with my actual smart contract, just my JSON Parser. I’m starting to think LINK holders are just trying to bullshit their way through technical explainations. Nothing you said makes any sense.

>>15717410
>I’m starting to think LINK holders are just trying to bullshit their way through technical explainations.
DELET THIS

>>15717233
If you have 5 data sources do you still need 100 Chainlink nodes querying them when you can just aggregate this obviously already decentralized data?

>>15717451
I don't know, do you?

>>15717466
Yes if you want to enrich Sergey and to introduce unnecessary risk to a contract execution.

>>15717451
100 was just an example. Probably want at least 2 per region/continent, using different isp/hosts, per source.

If you want to *almost* guarantee data gets into your billion dollar financial instruments, this is what you do. Your tokenized anime figurine trading dex won't need the contingency plans required, by law, for banks to have. So don't worry about it.

In the meantime, the solution for transacting millions of dollars a second will be an audited, legally enforced, and industry standard CL. Paper contracts cost businesses tens of thousands of dollars per contract per annum.

I realize this is just trannys trannying...but comon guy. Act like you've ever been a contributing member of society and can understand the basic economics and costs associated with running a company.

>>15717284
I've been wondering about this too. And if smart contracts take off and see a lot of use, won't api providers just start offering their data directly on the blockchain? Why connect link nodes to the weather station's api if the weather station already has a smartcontract you can pay and query for the data?

>>15717410
>I’m starting to think LINK holders are just trying to bullshit their way through technical explainations. Nothing you said makes any sense.
How are you just realizing this now?

Ok so why does it need a token though? Couldn't this just be an ETH patch?

>>15716963
>Ask synthetix what "lel just use a single public api" does the first time it glitches or inputs unsanitized data. (They lost 37000000 eth ).
chainlink doesn't fix that.
all cl nodes will just end out using that public api anyways. nobody will actually bother to do their own research when public sources are freely avaible.
oh, and the system actively discourages creating your own api due to the consensus method.

>>15717719
Correct me if I am wrong but it is more about Banks, forex and stock exchanges query market data. The transactions between finacial institutions is the only real use case I can see for this, if banks lose when there is a 0.001% incorrect price, this would make them more secure, automate even more and be used to increase bottom line.

It is for communication between very large companies and companies to government. Not some shitty weather station to fill a automated contract for a hail storm fucking up a bunch of normies paint jobs.

>>15716850
>use public data direct from source
>same data you pay the nodes for, no extra scam risk
I guess the problem with this is: the ones who use your smart contract then have to worry about YOU scamming them. And that isn't a problem with chainlink.

>>15717786
Alright then replace 'weather station' with bloomberg or something. Once bloomberg makes a smartcontract that can be queried for data (which most likely would happen long before the weather station anyway), why go through the extra step of using link?

>>15717905
More like why would ever single company bloomberg does business with go through the extra step.

>>15716850
i used it to become gay

>>15717804
A smart contract configured to automatically execute according to parameters from a data source would seem to be less risky than having a third party intermediate relay the data. Chainlink, as I understand it, seeks to migate this risk of using the 3:d party intermediate by giving you the opportunity to use several nodes’ data... thereby mitigating a problem witch arose from not using the source data in the first place. Maybe it also reduces the risk of incorrect reading of the source data by having several nodes... but has that ever been an actual problem? Is the issue something else, like that you have to use third party oracles/nodes to ever be able to execute a smart contract?

>>15717905
Staking. You can specify oracles to effectively insure/guarantee the data SC requestors ask for.

Off chain computation, and tons of extras that come with it. (Tees, threshold signatures/multi sig tricks)

Outputs to legacy systems and/or the legwork to integrate those systems back into your decentralized furry message board.

There's some real tech behind it, but CL has never been about that. It's just a marketplace/idea to bring node operators and people needing their data together. Unifying the payments. And making it much much easier for boomers to make the next step up. It's like being the sole Web dev in the early 90's.

Yes, technically you could spend years to reinvent what CL already has done, to save (less than the cost of a few years of all your programmers). There are tens of other companies that claimed they were going to do the same thing, but better. None of them still exist, or literally have bent the knee and are "partnered" to be customers of CL. If that doesn't say enough, you're retarded.

>>15718112

>Yes, technically you could spend years to reinvent what CL already has done
Would it really be that complicated though? Couldn't just the data provider program a smart contract that offers to pull data from its own api for a fee and the cost of gas? Why use the link token at all? And if I want to aggregate data from several data providers, can't be that hard to program in solidity can it?

>Off chain computation, and tons of extras that come with it. (Tees, threshold signatures/multi sig tricks)
This I agree with though, as a big plus for chainlink.

>>15718060
you're stuck in the 2017 eth mindset.
the smart contracts of the future won't be deployed by anonymous shady russians, they'll be deployed by bayer and dhl.
these billion-dollar companies won't risk getting sued by knowingly providing false data for financial gain.

>>15718326
It works until someone spoofs the data feed and clean the contract dry. That’s why you don’t use a centralized api

>>15716850
pic related...

smart contract usage won't be for shit like your biweekly paycheque (yet). the people who want smart contracts are the banks and whales, who are dealing with hundreds of millions, if not billions of dollars. they don't care about minor fees. they want security above all else, in situations where the counterparty would normally have incentive to fuck you over (eg credit default swaps)

I'm getting 2017 vibes ITT.

We're unironically going to make it frens. Ignore all tranny larpers with dates, conferences are nothing burgers 90% of the time. The only thing that matters in connectivity. Network weight and api adapters and market places for people to interact easily with this stuff. It needs to be simple like open law templates. Open laws stack with them link and eth is slept on but bullish af.

>>15718326
Sure, a data provider could employ programmers, sys admin/networking, pay for hosting, marketing, etc etc....or they can reply "k" to an email from a fat Russian saying "yo, how would you like to monetize data you're giving away for free...at no cost and no work to yourselves."

This is ignoring the decentralization and other (imo) fringe benefits that a single standard network/protocol would convey

Again, this isn't insane tech to get close (ignoring white paper shit like mixicles, tees, the node communication digisomthing)...it's WordPress for blockchain.

Hey boomer retard, welcome to 2020...take 5 minutes to chose some nodes from our 3/4/5 star rankings, customize a premade smart contract that's had the 50k in lawyers it would take you to do handled by openlaw, and bam...you just saved 50-90% of your business expenditures.

It's the package: the ease of use/deployment, and the legal/security all rolled up into a semi digestable package that can easily be sold as "we can make your company 10, 100, 1000 million dollars per year forever...it can be up and running end of business day".

That's a little bit more powerful and attractive to real world adoption than a fucking collectable kitty app or a way to track your coffee shipment from africa.

>>15719674
Fudson was bought on board to entice the furries

LINK is the only REAL project in the cryptoshpere.

The LAW of attraction is REAL.

Jesus was RIGHT (WING), NEETS will rise up and inherent the Earth.

We do not forget.

>>15716850
>implying this was ever about monetizing public data
>he doesn’t know it’s about monetizing for the first time swathes of private data

>>15719703
Goddam I hate the fucking trannyfaggot eth commies so much. It'll be a good day if one of the chink scams can actually deliver (jizzem feet, or a real scalable blockchain).

>>15719745
>jizemfeet pleeeeeeeease!

>>15719716
What does that even mean?

>>15719801
There's tons of data that isn't available, or will be quickly developed once it is profitable to do so. Soi are throwing a fit over btc/ltc pairs (which are being offered free btw). The reality this will incentivise advancement in areas like GIS https://en.m.wikipedia.org/wiki/Esri and other mapping. Iot data. Sensors.

There will be a newly blossoming sector just to diversify some of the single data sources. A literal gold rush to mine data from prospectors willing to track ocean currents (shipping/fuel surcharge settlements), better traffic, weather, medical, car/boat/rv accidents. Tons of stuff.

No one wants to do this now because it's all jerkoff science research or just not useful...but when it's saving a company 25k a year per contract X thousands of contracts...There's going to be a legit reward/bounty to scan and report the vehicle vins in an accident to trigger the insurance payout (or more likely be one of the many condition triggers with a body shop esig, etc to limit or eliminate fraudulent claims).

This is still really new stuff, and not going to filter down into everyday normie life until much later after the big stuff is refined and milked for all its profits. That can be easily done today, to the tune of multiple percentages of international GDP.

You niggers don't understand how slow trade financing/payments take, the response time, the litigation, and the end to end paperwork (plus fat donothing office npcs to rubber stamp) business takes. This is telegraph vs Google fiber, but for quadrillion dollar industries.

Complaining about paying 5 cents for weather data for your grindr gay sex dex to predict rain ruining your park bathroom hookup can wait.

>>15720034
Bump because Im phone posting and that wasn't worth it

>>15717284
>oracle services will grow in size but API services won't. the demand for decentralization won't create other competition on the API side

>>15720034
And Chainlink would be making this happen how?

>>15720677
By providing an incentive (money) if there's a demand for that data. Something that companies like google have struggled to do for their entire existence. That's also what CL is hiring people for, to develop oracles/data feeds...and what Oracle is doing for free to jumpstart this new ecosystem.

Again, all of that is on the back burner and is a very small portion of the somewhat basic functionality and feeds required to do billions in derivatives. Bankers don't need weather data or rfid tags, they need security/repeatability/trustlessness. In a tight package they can trust from someone that isn't a weird looking chink.

You redditretards are always wanting to jump to step 30, instead of 2. Always down into "but what if" details of hypotheticals instead of actual use, and basic functionality.

>mixicles, town crier, industry standard formatting (iso2202, gpi)
BUT WHAT ABOUT MY DECENTRALIZED DISTRIBUTED REMOTE DRAGON DILDO CONTROLLER BASED ON THE MAGNETIC WAVES OF JUPITER. FUCKING SCAM COFFEECOIN 2K EOY.

>>15721014
kek, learn some programming. Data providers don't need Chainlink, or even a blockchain to get paid.

Google already has countless paid APIs they earn money for. Literally every company in the world has.

LEARN SOME PROGRAMMING BRAINLET. YOU DONT KNOW WTF YOURE TALKING ABOUT.

>>15717410
So you propose using a separate entity outside the smart contract to retrieve and parse data... good idea! We can call that an oracle. And you propose to improve upon chainlink's design by... removing decentralization in the oracle layer and relying on a single static home-made piece of code to trigger your smart contract rather than a collateralized, industry standard, heavily audited, decentralized network of oracle services. Wow anon, it's almost like you're not arguing in good faith.

>>15721167
This. I'm a tcp ip developer at a fortune 50 big data company with 25 years experience in database infosec, I've looked into chainlink and all I can tell you if that the cryptographic measures implemented in regards to the decentralized security paradigm in the API and IoT structure of chainlink's github code is fundamentally flawed after the Pivotal tracker server crashed due to the core attacks on the network enabled by its corrupt data inputs and outputs, what this really means is that by attempting to solve the oracle sybil resistance issue it instead allows customers to bypass the encrypted hardware and even hack into the smart contract Intel SGX mainframe, unless they manage to increase the signatures and scalable nodes, which isn't likely considering the Google backend isn't compatible with the legacy JSON systems and Solidity language from the EVM in the Truffle stacks, that's why the ic3 and SWIFT engineering teams developed the ISO 20220 standards but it's centralized and susceptible to the 51% front running program so yeah basically Sergey didn't foresee that the enterprise customers and cloud blockchain dapps would never allow their protocols to rely on these permissioned host mechanisms thus rendering the LINK ERC 677 token obsolete and no serious developer would consider DLT technology in these conditions, sorry linkies I'm just telling it how it is.

>>15717451
100 is an arbitrary number but yes you need decentralized oracles to trigger a smart contract of any decent value, regardless of whether the contract uses a single source of data or an aggregation of multiple. Lots of reasons including most fundamentally... the oracle problem. How do you envision getting this outside data onto your blockchain if the miners/validators securing that chain all need to be able to consistently, independently verify the data used on chain? You cant have the ethereum network split in half for your shitty contract because all nine thousand nodes couldnt reach perfect consensus on the precise value of messy real world data (9,000 nodes x 5 data sources = 40,000 api calls that all need to execute perfectly and simultaneously every single time, for as long as your smart contract exists). Its just not viable. So we offload the data retrieval task to a dedicated set of specialized validators in a separate network aka an oracle network. Why use more than 1 oracle? Even if we assume your single oracle is perfectly benevolent and responsible (can never be bribed, coerced, or neglected), use of a single oracle introduces a single point of failure (a target for hackers) into what is fundamentally supposed to be an extremely resilient, tamperproof system. It just makes no sense to do. This logic does not change in cases where the contract utilizes just a single source of data. The oracle service still needs to be decentralized in order to reduce the attack surface for your connected smart contract (in this example limiting the effective attack surface to just the data source, rather than the data source and oracle).

>>15721306
Pasta.

>>15719068
Yes but like other anon said, that problem remains with chainlink and aggregation isn't that difficult.

>>15717451
Euronext is ONE data source
FTSE is ONE data source
etc

>>15721306
My concern is its primarily developed ad run on fucking linux

>>15721249
Obviously they aren't, all of the 'technical flaws' argued here only serve the opposite of what they were intending, subtle FUD.

The thing is, this has been a great thread, reminds me of old ETH days where the tech was actually discussed and not just money skelly with tubs of whey memes or fat sergey with big mac memes, pretty awesome!