[ 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: 446 KB, 1000x1000, China-Temporary-Construction-Galvanized-Chain-Link-Fence.jpg [View same] [iqdb] [saucenao] [google]
10021692 No.10021692 [Reply] [Original]

can anyone who understands the give me a run down on how I'd be able to implement Chainlinks smartcontracts as a company owner?

>> No.10021699

You don't theres no network, not even a test net.

its a meme coin.

>> No.10021703

Can you give me an example of your use case? You would only need chainlink if you needed smart contracts.

>> No.10021704

>>10021692
sigh...
OP you have finally come to the realization...
there are no actual use-cases for this shit in reality

>> No.10021707

>>10021692
I unironically would suck Sergey's dick 100s of times if he would let me

>> No.10021709
File: 9 KB, 250x243, You heard about it, right.jpg [View same] [iqdb] [saucenao] [google]
10021709

>>10021692
>He's not going to make it

>> No.10021711

employee payroll via smart contracts
supply chain tracking

>> No.10021720

>>10021692
what kind of company do you have? And do you have a use from trustless agreements?

>> No.10021723

do i need my own blockchain?

>> No.10021736
File: 191 KB, 804x922, fatsergeylink.png [View same] [iqdb] [saucenao] [google]
10021736

>>10021692

>> No.10021749

>>10021723
there will likely be services small businesses can get access to

>> No.10021771

>>10021703
>>10021711
>>10021720

I work in a grain trading company. We have contracts to execute from the purchase off a farmer, the contract between logistics companies to transport and the contract with end users such as feedlots, wheat millers etc

What do i need to implement these smart contracts - because at the moment a lot of our contracts are executed by paper and take up to two weeks to process

>> No.10021797

>>10021771
Wait, are you serious? Send a message to this guy https://www.reddit.com/user/vornth

>> No.10021813

>>10021771
You first need smart conrracts before using an oracle

>> No.10021824

>>10021692
depends on your positioning within the system. are you a data provider (eg an existing, external web resource), or want to pull in external data for your own smart contracts?

as a data provider, here's a very simple example of how you could use an external adapter to fetch data from your database (in this case, movie data) which is then used by an oracle in a job it runs (eg compiling movie data from various sources into a detailed catalog)
>https://github.com/thodges-gh/ExternalAdapterExample/blob/master/src/main.go

unfortunately, the only core component of the CL infrastructure currently open-sourced is the oracle node. we will need to see how aggregation and reputation work to get a full picture of how you would go about taking in these multiple data sources and outputting something useful.

>> No.10021860

>>10021771
Have you posted before? I think I remember something about a farmer wanting to use smart contracts to ensure the cattle he was going to buy were fed properly or something like that.

An example smart contract for you, if I'm not mistaken, could look something like this:

You buy some bushels of wheat from farmer Bill and use the transport company swift to ship them to you. Swift guarantees delivery of said grain by July 01. You could use a smart contract to ensure payment penalties for each day past July 01 Swift takes to deliver, thus ensuring you pay a fair price for delivery.

Another example could be that you are guarunteed by farmer Bill X bushels of wheat. He (or swift I guess) recieves a payment penalty for each bushel that you don't receive in good health.

>> No.10021864

>>10021771
Here's the beauty of the Chainlink approach:

Once the full suites are out, you'll just have to write the same native language contracts you write now and let Chainlink, ETH and SWIFT do all of the heavy lifting in the background.

From a realistic standpoint I would do the following:

- Tell whoever does your contract work that you want to run a small number of low value contracts as a test when this comes out.
- Tell them to familiarize themzelves with Zeppelin_OS and what smart contracts are
- Wait ~6 months or so until non-launch partner Smart contract demos are live. At that time Smartcontract.com and Zeppelin_OS will have an army of adoption specialists ready and waiting to help you

>> No.10021873

>>10021813
>>10021813
for now, but in the end this will not be entirely necessary. eventually there will be entire companies built just to make Chainlink contracts as easy to create as drag-n-drop and filling out a form or two. Just specify the data sources you're working with, the ways in which you can change it (state transitions) and which parts of the data the oracle needs to use/transform

>> No.10021894

>>10021771
APIs are not intelligent enough today even with machine learning to decipher whether the grains are correct kind, whether they are still in good condition, or even if they are delivered. If you were to use smart contracts today for transportation/delivery you would get rekt by bad inputs from unreliable sources

>> No.10021924

>>10021894
this

>> No.10021941

>>10021894
this is where sensor data comes in. you could use cameras to get visual imagery of the product, and bio-analysis equipment to verify that the grain has the correct chemical composition.

It doesn't make sense to think of an API as intelligent or not intelligent - it's just a gateway for whatever intelligence you have. With enough data, any ML model(s) could be deemed sufficiently intelligent

>> No.10021999

>>10021894
I was thinking about this. I don't think having a human inspecting the grain and inputting data manually into the smart contract removes all of it's utility. When done this way there is no need to trust lawyers or middlemen that appear in contract litigation - just trusting farmer brown. Not to mention how much faster it all could move.

>> No.10022013

>>10021941
And where would those sensors go? What is keeping anyone from finding those sensors and fucking them up? The issue with that is that the location of these things can’t be decentralized - they HAVE to be near or around the grains to work at all, and if you want more inputs for redundancy they are going to be easier to find. The closest thing we can get to measuring physical phenomena in a decentralized manner is the local weather because the sensors could be anywhere.

>> No.10022054

>>10021999
Trips, checked
If that's the way it's already done, then yes using a digital interface to track instead of pencil and paper would speed things along. No need for a smart contract or chainlink though

>> No.10022059

A blockchain is a hermetically-closed virtual world. It can only contain claims that prove things about itself.

For example, a transaction may contain proof that the private key associated to the address mentioned in a previous transaction has signed this new transaction.

It is even pointless to try to use a blockchain to prove anything about the real world, because proving anything about the real world is fundamentally impossible.

Some people simply refuse to graduate to the understanding that proof is only possible for an abstract world for which you can provide the complete construction logic. You cannot provide the construction logic of the real world, because you did not build the real world by yourself.

For example, a blockchain is absolutely not usable for KYC ("Know Your Customer") or AML ("Anti Money Laundry").

In the real world you cannot prove that John is John. He may say that he is John, and he could provide some paperwork that seems to say that too, but all of that does not prove anything.

President Obama could not prove that he was born in Hawai, if only, because such proof cannot possibly exist. Imagine that you personally saw a baby in the hospital back then. On what grounds can you insist today that it is the same Obama?

>> No.10022065

>>10021692
you're an idiot. nobody uses smart contracts, therefore chainlink is useless. you're about 10 years early faggot. chainlink is a piece of shit meme coin, and you're a pleb for buying it

>> No.10022076

>>10021999
I wonder what Sergey would think about this. It’s not end to end trustlessness like he says he hopes to achieve. If you have to trust any single party with contract input I doubt he would endorse using it that way

>> No.10022126
File: 46 KB, 264x678, Salesforce.jpg [View same] [iqdb] [saucenao] [google]
10022126

>>10021771
get in contact with salesforce.

>> No.10022149

>>10022059
Well shit. Then we'll need to move humans into the blockchain.

>> No.10022216
File: 56 KB, 642x224, wheat.jpg [View same] [iqdb] [saucenao] [google]
10022216

>>10021860
>>10021894
>>10021999

I've never posted here before. But I can paint a picture for you on the current operations. Farmer X harvests grain and he then decides based off of the quality prediction (given by the technology in a harvester) where he sends the grain. It can be to a feedlot if it is the poorest of quality (FEED) and from there it goes up in price as to the quality for bread making etc. Some years FEED wheat can be worth the same as Prime Hard 1 (APH1) due to the demand from feedlots when cattle prices are high.
When feedlots purchase wheat quality isn't something that is of huge importance because the only thing they are looking for is the cheapest form of energy to feed the cattle.
In terms of the proof of quality product it essentially is just a tick of a box from the feedlot receiving the grain to say theres no contamination from insects such as weavels etc.
If I were to purchase or gain access to 15 different API's regarding price of FEED wheat in a specific area + also price data on transporting costs - is it possible to feed this data into a contract? So that live time I can be offering the best going rate always for a feedlot?

>> No.10022243

>>10021860
Whats wrong with using a conventional contract in these scenarios?

>> No.10022267

>>10022243
OP said they were too slow, taking up to 2 weeks to process

>> No.10022339

>>10022216
The risk of using a smart contract boils down to the fact that it is irreversible and executed automatically. So knowing that, you have to ask yourself if you are comfortable trusting whatever parties are responsible for the inputs for that contract (status of delivery, pricing, etc) and whether you can afford a botched input every once in a while. The only thing that chainlink will be sure to automate and save time on is obtaining price input, given enough API sources and nodes. So whether you choose to use smart contracts or not, chainlink can still save you a little time if your current protocol for checking prices is by manually checking multiple sources.