To prevent Sybil attacks Chainlink will utilise off-chain reputation providers to validate node operator identities...

So what earns your node reputation? Is it just the amount of up-time and consistency of accurate information or are there other factors?

Very little info out at the moment. Yes to the things you listed, and apparently tokens held in the node will have some bearing as well.

Will the node operator be completely blind as to what kind of information his node is processing, or will each node be married to a specific API/datasource?

My understanding is that it will be incumbent on the node operator to seek out specific data sources and put adapters on them. So node operators will have to actively compete with each other to find the most lucrative data sources. I expect this will be a whole other area where a data marketplace will come into existence to facilitate node operation.

Hmm that is interesting. Doesn't that effectively mean you'd have to be a data provider and running your own node for you to really make any money?

yeah don't worry all you faggot anons have all sorts of useful unique data to share with the network
we're all gonna make it bros

Well, one of the arguments for the token value is that most of the tokens will be locked into nodes, which will reduce supply and drive up the price. So I'm not really sure how this here is bad in that regard.
As for running nodes: attaching yourself to a company's API that provides free, publicly available information should still bring in some cash (it has to be incentivized or there won't be nodes to fetch this info for smart contracts that need it). But, it seems that having a monopoly on a useful information source is where the real money is (i.e. companies providing this type of information stand to gain the most). I'm not really sure how a neet would position himself to take advantage of any of this.

if you're attaching to a free and public API then there is no barrier to entry other than an internet connection which even an african can get
sounds like you'll be making a lot.
locking up tokens magically make inflows and price rise sir? good invest. so it's basically no better than a masternode scamcoin tokenomically?

>locking up tokens magically make inflows and price rise sir?
Effectively reducing supply will increase the price if demand stays constant. Dunno how this will work in practice with Link

>so it's basically no better than a masternode scamcoin tokenomically?

Basically what I'm trying to find out here.

>once we start getting these contracts developed
>once we start
>>>start
we're going to be holding these bags for years arent we

Attached: 1525863808273.png (381x353, 40K)

You're gonna carry that weight

Attached: IMG_20180528_220850_350.jpg (960x696, 55K)

No fuckjng way. Even if I knew how to operate nodes I'd be fucking out once the coin skyrockets in price

I'll take the lump sum powerball payout thank you very much. I have much more important things to do then run nodes... like planting bags off the side of intereste highways in the middle of nkwhere with 3k dollars in each of them and posting threads on here where the coordinates are so no linkers can have my scraps

Yes. I think a lot of people here are far too optimistic about getting rich off link without actually doing anything. If you want to definitely get rich off link then run nodes. That's basically the crux of my OP.

You’ve got to make it harder

Yep, seems that way. What I'm not clear on is how the economics of various data sources would look like.
Can an unlimited number of nodes attach themselves to a single data source? Let's say this happens (well, not unlimited but arbitrarily large). A smart contract user/creator might specify that they want 10 nodes to validate the data. Will the 10 nodes with the best reputation automatically be selected to provide it?
If so, and if the numbers of tokens held are one of the factors determining reputation, wouldn't that give an oligopoly rich node operators over say.. Jow Forums users?

Yes. Is that a problem? I would say a corporation holding a million link is probably more trustworthy than a racist neckbeard neet from Jow Forums more often than not

>Can an unlimited number of nodes attach themselves to a single data source?
Yup
> A smart contract user/creator might specify that they want 10 nodes to validate the data. Will the 10 nodes with the best reputation automatically be selected to provide it?
No, all the criteria will be customisable, includign reputation. You can set minimum benchmarks and then that will reduce the pool to any nodes that meet the benchmarks, but after that node selection will be randomised.
>If so, and if the numbers of tokens held are one of the factors determining reputation, wouldn't that give an oligopoly rich node operators over say.. Jow Forums users?
Yes and no. It has been suggested that holding over 20k link in a node will have diminishing returns in terms of reputation. For the absolutely apex high value contracts they might demand a reputation score so high that any node with less than 20k won't be selected, which will preference wealthy node operators, but that will only relate to small class of the total number of contracts using the network. It makes sense, anyway, because the high value contracts are probably going to need higher amounts for penalty staking, so will be better suited to the wealthier nodes anyway.
Broader questions about whether it will be worthwhile for your average 4channer to run a node in competition with these highly resourced entities can't be answered yet, obviously, but I would say that Chainlink wants a diverse node ecosystem. It's in their interests to have a large number of nodes. Better redundancy, better network effects. So I would say that people who get on board at the ground level and run nodes from scratch should do very well, even without the resources to secure the top level contracts.

>Will the 10 nodes with the best reputation automatically be selected to provide it?
It seems like node operators will be able to charge what they want. Obviously, the better your reputation, the more you can charge and the more jobs u'll get. A competitive node market is important to establish reputation. So it won't be completely random.

>Yes. Is that a problem?
Depends on your perspective. I'm just looking to get a base understanding of the system here.

Thanks, that clears it up a little. Is there any place that does a good job summarizing what is currently known about the way nodes work (conceptually rather than technically)?

Not really, because the reputation and consensus code is not publicly available yet, let alone any actual guidance or published material about the structure of the ecosystem, incentives, governance, etc etc. Prospective node operators are (mostly) just sitting on their hands at the moment. Hopefully within the next few months things will become clearer.

>Even if I knew how to operate nodes I'd be fucking out once the coin skyrockets in price

Thats what you dont get, the price is only gonna moon AFTER nodes are operating, because holders will start getting a hefty sum for their services, much higher than the price of link will be, and they will transmit that information to the market, by buying more tokens, which will make the price skyrocket.

But before that, it just seems unlikely that link will be over $10, even if we are back in a bullrun.

Exactly. Token demand is driven by node operators. Demand for nodes will only increase when people see how much money node operators are making. But because early node operators will be getting paid (abundantly) in link, and THEN link will moon, it's the node operators who are poised to make far more than holders who simply have to sit around and wait for the network to establish itself.

It takes 20k to be a node operator, correct?

No. A node can have any amount of link in it, but more tokens will contribute to a higher reputation score. 20k is seen as the point past which there are diminishing returns for staking more tokens.

>Token demand is driven by node operators
I assume for this to be profitable and viable, the cost of fetching API data would have to be fairly low (expensive fetching would severely limit the use cases I think), but a node operator could still get a good amount of income by processing a lot of throughput. Does that sound about right?
If so, should we expect node income (and token value for that matter) to start off relatively low and steadily rise as adoption (demand) increases?

Is it known if bandwidth-heavy vs light data requests have any effect?

The node software so far is incredibly lightweight and the actual data/bandwidth usage is not expected to be in any way strenuous.
Node income will scale with the network, so yeah, I wouldn't expect any substantial increase in the link token price until node operators are doing considerable work. The good news is, I wouldn't be surprised if Chainlink hit the ground running and had quite a lot of network use almost out of the gate.

Cant you just run multiple 20k staked link nodes?

Only thing Im worrying about is if I need to invest in some kind of new rig for the node, Im not very tech savvy, and I would be scared of the power going down and losing my links.

I would think so, but if you look at Thomas' post in the OP the identity/reputation system may actually limit the number of nodes a single entity can operate. That's not a definitive answer, though.
You wouldn't run it on hardware at home, you'd run it on a VPS.

Are you concerned about the scaling issues of the Ethereum network? I imagine that truly wide adoption would likely require something on the scale of millions of transactions per second.

I know the Link token is blockain agnostic. If Ethereum fails to scale sufficiently, does it seem likely the token will migrate to another platform or operate on multiple platforms simultaneously?
If so, would a node operator be able to run only on one Link-utilizing platform or could a node handle requests for multiple blockchains at the same time?

So why will the token not increase in value

>what is speculation

The data providers will either seek node operators to monetize their data, or seek tokens to run their own node

>So I would say that people who get on board at the ground level and run nodes from scratch should do very well, even without the resources to secure the top level contracts.

This. Aim to service smart contracts that need regular and accurate data. Lets say an insurance contract - so, weather is a good example. Spin up a test node, and start working on consuming some weather data from a reputable source. Get ahead of the vast array of potential use cases.

boards.Jow Forums.org/biz/thread/9655412#top

check this thread out some weird shit going on

kill yourself

The EVM is only used to handle payments, specifically the movement of link tokens. I think Ethereum is more than capable of this, and if by some calamity it couldn't handle the volume then migrating link tokens to a different network wouldn't be impossible.

How would that look in the case of a higher frequency data feed (e.g. constantly changing market data)? Is a Link token payment made for every price update?

There are 8 factors, the info is literally out there. But most of you are too lazy to look for it.
Reputation depends on factors like uptime, response time, tokens staked, etc.

The node operator will be blind to the data only if he's running a node with intel sgx, which will most likely be required for privacy intensive data fetching (bank contracts, etc. )

>more tokens will contribute to a higher reputation score.

It depends on the contract creator. There are multiple factors on the reputation of a node.
Some contract creators might want to not take the tokens staked into account for the reputation.

Yes, but it won't guarantee that your node will get chosen to fetch data.

Again, I imagine this is all customisable, but that would be one way to do it. If there are thousands of API calls over, say, a week, though, it wouldn't make sense to demand payment on a "per call" basis because you'd be copping so many Ethereum fees.