Gmx v2: new low-latency chainlink feeds

[William88]

Variable fees are for service providers who:

1. Take financial risks to provide services (liquidity providers, financial intermediaries)
2. Directly refer users or bring in revenue

Giving a variable percentage fees to an oracle provider that does neither of these things sets a horrible precedent for GMX and all of DeFi. Next we’ll have eth.limo charging percentage of revenue for hosting fees? It’s preposterous. Your eCommerce masters charge retailers a percentage of revenue because they are directly referring customers to them or have a monopoly to act as an expensive payment intermediary.

[gmx-advocate]

Chainlink as a company might be legit, but as a defi project is a complete let down. they released a whitepaper back in 2017 in which they promised on delivering a completely decentralized network of nodes that would provide smart contracts with completely decentralized and manipulation proof data.

those were all promises, and even though they released a whitepaper v2 in which they entered into more details regarding how they’d manage decentralization. IN THE END, after 6 years of developpement and USING THEIR OWN PREMINTED TOKEN to finance their COMPANY.

chainlink has proven to be a fully centralized oracle network fully in control by the company.

Chainlink labs has no intention on becoming a decentralized network, they have whitelisted a few enterprise grades nodes ( the whitelisting process is completely opaque ) and chainlink labs maintains those nodes on behalf of those enterprise. NONE OF THE NODES THAT RELAY THE DATA TO SMART CONTRACTS ARE INDEPENDENT ENTITIES, THEY ARE ALL IN COLLUSION WITH CHAINLINK LABS and have common financial interests.

not only the chainlink network data source is opaque, but the nodes relaying the data are in collusion with the foundation, also, they have a poor security regarding how the multisig is managed a 4 out of 11 and basically we GMX would be putting our whole exchange between the TRUSTY hands of those 11 individuals.

the chainlink company is an opaque organisation, that has been “subsidizing” their network for 6 years using their investors money to pay for an expensive hiring strategy of talents that resulted in not delivering on any promises made to their token holders / investors.

chainlink also constantly gaslights their community into believing new d3velopments xill come up soon to only deliver alpha profucts in v0.01.

key points are

• as long as the chainlink node operators are whitelisted by chainlink labs (permissioned network) we are at risk of sybil attacks

• the chainlink team has a track record of selling their token to fund their company and enrich themselves, they would use our token to crash our market cap

• the chainlink team has encouraged many illicit initiatives like celsius, linkpool and bancor, they even colluded with ftx and sam bankmanfried

• the chainlink team, has had previously defected with their price feeds reporting silver prices instead of gold, or staggered delayed wrong average data coming from cmc

i strongly advise against using the chainlink centralized oracle (that promotes itself as decentralized) and hope that gmx finds a better solution, than use that scam.

[Pochacco]

So to get this straight: you believe GMX should go with Chainlink, but only for a fixed fee? What would be a reasonable fee in your opinion? How do you put value on securing a network/platform that is making millions in user fees, besides as a percentage fee? Whatever number you are going to come up with, it is going to look stupid whenever the platform expands or contracts. (yes, it works the other way around too. If GMX collapses, the fee will drop accordingly).

Your eCommerce masters charge retailers a percentage of revenue because they are directly referring customers to them or have a monopoly to act as an expensive payment intermediary.

Neither. It is a convenience thing. If a certain process in your business is very time consuming, high risk or specialized, outsourcing for a small fee is an obvious choice. SaaS took off because doing everything in-house is a huge headache.

Consuming safe, low latency oracles is what eventually generates revenue for GMX. They could do it without an oracle service, but as shown by the billions that have been hacked in the past few years, that’s extremely short sighted and will end up with their platform exploited. So, instead they opt to use an external, specialized service that promises to not only provide Oracles, but attach extra guarantees about latency and uptime as well. In return, this party asks a small % fee for their efforts, incentivizing them further to actually want GMX to do better, so they can profit as well.

Extremely reasonable, and purely by my own anecdotal evidence an arrangement like this works miles better for both parties than a fixed, flat fee since there is incentive for both parties to grow and innovate together.

[gmx-advocate]

Chainlink, while being a legitimate company, has not met the expectations of the DeFi community as a decentralized project. In their 2017 whitepaper, Chainlink committed to delivering a completely decentralized network of nodes to provide smart contracts with tamper-proof and decentralized data. However, after six years of development and utilizing their pre-minted token to finance their company, Chainlink Labs has not followed through on their promises.

Chainlink Labs has no intention of becoming a fully decentralized network and has instead whitelisted a few enterprise-grade nodes, the whitelisting process being opaque. Chainlink Labs maintains these nodes on behalf of the enterprises, and none of the nodes relaying data to smart contracts are independent entities; they are all in collusion with Chainlink Labs.

The opacity of the Chainlink network’s data source, along with the lack of independence of the nodes relaying the data, raises concerns regarding the security of the multisig, which has a poor rating of 4 out of 11. Entrusting our exchange to the hands of these 11 individuals is a significant risk.

Chainlink’s company has been opaque and has used their investors’ money for six years to finance an expensive talent acquisition strategy without delivering on their promises to their token holders. It is essential to note that as long as the Chainlink node operators remain whitelisted by Chainlink Labs, we remain at risk of Sybil attacks.

Furthermore, the Chainlink team has a history of selling their token to fund their company and enrich themselves. They have also been involved in illicit initiatives such as Celsius, Linkpool, Bancor, and colluded with FTX and Sam. In addition, Chainlink has had issues with price feeds reporting silver prices instead of gold in the past.

Based on the points mentioned above, it is strongly advised that GMX looks for a better solution than relying on a centralized oracle such as Chainlink.

[William88]

If Chainlink were actually decentralized as the original whitepaper laid out, there would be LINK staking with SLAs for node operators to compete for our business. Instead, we have a monopoly fudding competition and trying to to grift themselves 1.2% of the fees. Yes, fees should be charged per call, and node operators should compete for who will do it for the lowest cost. Probably 50-100% over operational costs and they’re rolling in fees. But chainlink is neither decentralized nor a fair marketplace. Node operators are whitelisted by Chainlink and most feeds aren’t even very decentralized. Imagine calling 7 node consensus dEceNtrAliZed lol!!!

[Pochacco]

To bootstrap the network, Chainlink intends to use node operator’s IRL reputation to make sure they don’t misbehave. This is a necessity before slashing mechanisms and reputation are live. Yes, this means Chainlink has to whitelist node operators themselves. This also means they can get away with an (admittedly) low number of nodes for now.

The other route would’ve been to go completely decentralized from the start as you say, and risk exploits and bad actors along the way.

I think there is something to say for both approaches. For now Chainlink’s approach seems to be working fine, even if it is far from decentralized. Personally I think they are demonstrating well how the network is able to function and eventually scale and generate profit. Whether they can actually translate that into the promised, fully fledged decentralized ecosystem including NEET nodes, is the million dollar question on whether you should continue to hold or sell.

[William88]

It works fine, but not a product that should be charged for. If feeds were decentralized and economically secured by node operator stakes, the percentage fee might be warranted. But if DeFi protocols bow to their revenue timeline this early on in the product development, its an admission by them that decentralizing these feeds with cryptoeconomics will never happen and we are basically accepting that. What if 2 years go by and there’s no improvement? Will we claw back years of payments we made for feeds that were supposed to be eventually decentralized?

[UpsideRunway]

@William88 can you confirm whether you are or are not the same William87 who is on the Synthetix Spartan Council?

Gauging from your fake cease and desist letter and your other adversarial comments, it seems like your goal is to create chaos in an attempt to cripple GMX’s future growth potential, as GMX is a direct competitor to Synthetix and their perps product. Competitors should compete on the basis on who has the better product and UX, not thinly veiled attempts to attack one another on governance forums.

Integrating Chainlink is obvious win in terms of improving the competitiveness of GMX. Paying Chainlink with a percentage of the fees generated by GMX means they’ll be economically incentivized to continue improving the oracles, so GMX can obtain more market share and generate more fees.

The proposal is for integrating Chainlink’s low latency oracles, but it’s also onboarding Chainlink as the oracle team of GMX and solving all future oracle needs. So, it’s pretty silly to state that the oracles should simply not be paid for, that is not a sustainable approach. Pyth will have to charge Synthetix as well if they want to be sustainable, can’t rely on Jump’s dirty money forever.

I also find your arguments around node whitelisting a bit ironic, given this applies just as equally to Pyth used by Synthetix, if not more considering it’s not disclosed who the Pyth nodes are on each feed. All oracles have a degree of trust required, but they’ll continue to improve over time, especially as there’s more resources to do so. Going from GMX’s currently centralized oracles to Chainlink networks (even if whitelisted) is a clear win in improving decentralization.

[Pochacco]

What if 2 years go by and there’s no improvement? Will we claw back years of payments we made for feeds that were supposed to be eventually decentralized?

Chainlink is providing low latency, decentralized oracles to users. To what degree you can call these oracles “decentralized” is debatable, but they are still providing users with low latency oracles. This is what they pay for. Why would they want their money back for 2 years of perfectly provided service? They pay for what Chainlink does NOW, not what we all hope the network is going to do in the future

Also answer the question

[UpsideRunway]

Hi @gmx-advocate, are you aware that edit history is public on governance forums?

I’m curious as to what your intentions are, given the level of hyperbole and misinformation you are providing here. Who asked you to write/edit/publish this comment? It’s clear you’re attempting to appear as a GMX community member “gmx-advocate”, but it’s obvious you are not. Rather than debating the merits of this proposal, it seems you’re only attempting to smear Chainlink’s reputation with misinformation. Getting some serious Zeus Capital vibes.

I’m not sure what your true intentions are, but one thing I’ll note is that I find it funny you brought up FTX/SBF, when SBF’s own published balance sheet of FTX shows an investment of $63,766,000 into Pyth.

[UpsideRunway]

Your comments are giving me heavy Zeus Capital vibes, where an avalanche of unbacked and hyperbolic claims are made, meant to overwhelm the reader. This is called the bullshit asymmetry principle, “the amount of energy needed to refute bullshit is an order of magnitude bigger than that needed to produce it.” You are aware that the burden of proof lies with the one making the claims, and not those refuting it?

But I digress, sure, let’s take a look at your claims.

chainlink node operators are selected nodes that collaborate together with chainlink labs

Anyone can run their own Chainlink node, it’s entirely permissionless. Anyone can also create their own oracle network on Chainlink composed of any nodes they want. The Chainlink team are experts with oracles, so yes they’ve launched oracle networks which act as a sort of managed service, where nodes are chosen by the team based on past performance and their reputation. Most developers are not experts in oracles and don’t want to manage their own oracle networks and handle choosing nodes, but I’m sure if the GMX team wanted specific nodes in the oracle networks, Chainlink could help with that too. Most oracle networks work this way.

chainlink labs has monitoring internal tools that monitor each node data and they alert them directly whenever there is a discrepancy, some nodes are even setup by chainlink teams themselves

Do you have any proof of the claim that “some nodes are even setup by chainlink teams themselves”? I’m not aware of any such nodes. For the other point, I don’t see the issue with the Chainlink team monitoring node performance? Such monitoring, along with each node’s monitoring setup, results in more reliable networks. This is pretty standard for oracles, I’d be concerned if Chainlink didn’t monitor the networks.

chainlink labs has no source of income beside the ICO funds and the tokens they dumped during these last 6 years ( they made different blogposts on it)

Do you have any proof of this claim and that other users don’t pay anything? Regardless, this seems hardly relevant to the proposal. Why do you think Chainlink is asking for a portion of GMX’s fees in exchange for integrating low latency oracles and onboarding the Chainlink team as GMX’s oracle partner? It’s to create sustainable oracle solutions for GMX so LINK subsidies are not required. That’s kinda the whole point of asking for fees to use the service.

chainlink price feeds did effectively have hiccups and misreporting due to unverified unauthentified unsigned sources of data the nops use

You mean the gold/sliver mixup that happened over three years ago? The tech stack of Chainlink feeds have significantly improved over the years and there haven’t been any notable issues since, even when Ethereum gas prices went insane during DeFi summer or the volatility during FTX collapse. It would be great if data providers signed their data, but I don’t see how this is on Chainlink? They’ve explained many times how the data aggregation process works for their feeds and has clearly been proven to work very effectively, where most oracles have repeatedly failed (Pyth).

none of the chainlink node operators reveals from where they fetch their data or how they compile the data they report on chain, it’s all “just trust us”

There was a blog linked above that explains how data is aggregated in Chainlink feeds. Here is another blog that explains in more detail how the data aggregation process takes place: How Chainlink Price Feeds Secure the DeFi Ecosystem | Chainlink Blog. Data is pulled from a number of premium aggregators like Kaiko, Tiingo, CoinGecko, CoinMarketCap, etc, whose entire business model is to create accurate prices for crypto assets. This is explained in those blogs.

chainlink has been claiming for years they build a decentralized oracle, so far i fail to see in what way their Dons are decentralized they even shutdown recently some price feeds because some of em only had a couple of node ops reporting

Chainlink isn’t a single oracle network, but a protocol that enables oracle networks to be created. It’s explained in the blogs above, but Chainlink networks are decentralized at the data source, node operator, and network levels. Each oracle network consists of multiple nodes, where each node pulls from multiple data providers, and each data provider compiles data from across multiple exchanges. Then the data is all aggregated together and published so dapps can consume it.

Some feeds were shut down, not because “some of em only had a couple of node ops reporting” but because there wasn’t enough demand for those feeds. Doesn’t make sense to subsidize a feed that nobody is using. It’s why fees are important for oracles to continue operating.

chainlink doesn’t have 1 fully released product, all their products are in prealpha alpha or beta, and even those in beta either are closed betas or have limited access

This is just a flat out provable lie (Zeus Capital vibes). Price feeds, proof of reserve, vrf, and automation are all products live in production and are used by countless dapps. Sure, improvements are made to those products over time, but why shouldn’t they be improved over time? They are still very much fully released products. This point isn’t even relevant to this proposal, given GMX would be using the new low latency oracle product that was developed together with GMX. Pure FUD.

chainlink node operators can easily be corrupted as many of them already run nodes for other blockchains or oracles.

What? This argument makes absolutely zero sense. Chainlink nodes have economic incentives to remain reliable and honest, because if they don’t, then they harm their reputation as a business, lose all future revenue they could have generated in the Chainlink network, and harm the value of the LINK they’re paid in. Feels like you’re really grasping for straws on this one.

chainlink the company is fully opaque, and has never released any roadmap or plans for improving their systems in regards to decentralization and collaterization

Chainlink just published a product update for Q1 recently, which details recent and upcoming developments: Chainlink Product Update: Q1 2023 | Chainlink Blog. They also published a roadmap for staking last year (ran out links, just google Chainlink staking roadmap). Just more baseless FUD here.

chainlink doesn’t explain how those gmx fees will be used or allocated, will they convert them to usd? is there somewhere a smart contract that makes sure everything is handled in a trustless matter? where will gmx fees go? directly into chainlink labs wallet? or in a sort of onchain treasury will they dump them instantly?

To cover the operating costs of the services used by GMX and to fund future research and development on improving those services I assume? With this proposal, a portion of the fees generated by GMX would be paid to Chainlink. GMX doesn’t generate fees in GMX coins, so fees wouldn’t be paid in GMX coins. It doesn’t really matter if the fees are converted to another asset or not, who cares if ETH or stablecoins are dumped?

I’m sure you’ll probably post another comment with an even larger laundry list of FUD points that would require even more effort for someone to refute, but I hope it’s clear by now that you’re not being genuine in your arguments. I’m a long time LINK holder, so I’m not completely unbiased, but I’ve also been a long time GMX user and staker (GLP too). I want what’s best for the GMX ecosystem, so it can generate even more fees for me to collect. Seems obvious that integrating Chainlink would help facilitate that. I’m still curious what the underlying motivation of “gmx-advocate” is for posting all this FUD?

[coinflipcanda]

The idea that GMX grows 10x and has 10 million dollars allocated towards our Oracles and data feeds is not a flaw in the system but a feature. At that stage the GMX ecosystem would be generating over 800m in annual fees, volumes and open interest have likely also scaled significantly. This means infrastructure needs go up but also more and more attack vectors especially with regards to Oracles open up, GMX isn’t the side story but the main focus of the effort of those actors at that stage.

You don’t want to rely on general public general infrastructure, you want to have the largest Oracle player in the space, focused on their biggest client and making sure that our safety, security and requirements are met because they are deeply vested in the continued success and proper functioning of the protocol.

As GMX scales the incentives for providing malicious data continue to increase as the potential payoff is higher. The higher the protocol fees in nominal amounts that we pay with consistency over time, the greater disincentive there is for data providers and node operators to submit incorrect information, the greater the investment in data security and working to ensure the oracles provide a true representation of market price not manipulated by data spoofing and the like, the less likely it is that these attempts succeed.

GMX doesn’t need just a simple service provider, it requires a partner deeply focused on the current and future needs of all things related to Oracles, data feeds and data security.

Just like we would never take our traders and liquidity providers for granted as they are a key part of making the protocol viable, Chainlink, their data providers and node operators have the same incentives for a growing GMX and the growth in fees that it could bring to them.

With making an allocation of 1.2% to support Oracles we are signalling that as a protocol we are committed to investing and re-investing over time in the quality of our data and its security. We would expect Chainlink would make the requisite investments to satisfy the community of the value they create for the protocol through the efficient functioning of our markets, and if this is not delivered the the DAO can make changes in its strategy and partner mix.

[Ikun]

First of all, the cooperation between gmx and chainlink is appreciated, and it is a win-win situation for both parties.
1, But I am worried about the 1.2% profit commission. May I ask how Chainlink calculates the 1.2% commission. Why not 0.6%, 0.8%. Why fixed cost pay per view.
2. What percentage of the commission can cover the cost of chainlink and keep it profitable?
3. Is the 1.2% profit commission, the result of the cooperation between gmx developers and chainlink, fair to gmx?

[relied]

Can we make this forum write-only for stakers? Maybe even give it the need to have a certain number of MPs. It´s tiresome reading through troll comments from people trying to promote a different dex perp. In all seriousness, this is the worst what you can do for your protocol. If you do think otherwise…well…you´re wrong.

Regarding the proposal: I am absolutely in favour of this. I can not comment on the reasonability if 1.2% is a “fair amount” or not and I doubt anybody here can give an answer to this. Chainlink for sure invested multiples of this out of theirpockets to develop a step-change solution for the dex perp sector. Since fees are proposed to come from V2 volume it is also speculative how fast this will scale / if it will scale at all to current levels. They´re taking a high risk here and should absolutely be rewarded for this accordingly.
Additionally, I see a great opportunity for the GMX protocol to have a strong partnership on the development side to an established team like chainlink. Implementing their oracle solution and using it for expansion to different chains / improving the underlying tec is an advantage that can not be undervalued.

I highly appreciate the innovation that is happening here and encourage people to see this as an opportunity and not to bargain if 1.2% or 1.0% is a fair return.

[dageda]

Your explanation is very good, and it is also very forward-looking. However, whether Chainlink can provide services that match the corresponding fees after receiving rewards, I think there should be stricter requirements on fees and requirements, and the percentage of commission should not be Gradually increase linearly, for example, it can be temporarily at 0.5%. It is important to know that the growth rate of the gmx protocol may turn 0.5% of the fee into a huge income. Therefore, if gmx is growing at a high speed, a stronger price feed is required mechanism, it means that 0.5% is also a high-speed growth fee, so it can also pay enough for the oracle machine. On the contrary, if gmx does not have high-speed growth expectations, the requirements for the oracle machine are not so urgent. The future development of the community requires more investment and expenditure, not just for the oracle machine. Can we leave more reserve allocation space for the community? Therefore, in order to deal with the problem more efficiently, it is recommended to adjust the commission distribution ratio of chainlink cooperation as soon as possible, so that things can achieve better results. In community discussions, we found that a simple commission of 1.2% is unconvincing. Thank you I hope that the gmx community can be more efficient and harmonious!

[Pochacco]

So much text that says absolutely nothing.

[coinflipcanda]

There are no doubt many new members of the governance forums who appear to be deeply committed to Chainlink or Synthetix. We welcome you into our community, but would ask that you keep your tone civil, stay to the topic of the discussion and if you have a formal roll in one of the other communities you share it so that it might provide context to your messages. All constructive feedback is appreciated.

@William88 i’ll pose my question to you again, if you are the william87 who sits on the Synthetix Spartan Council please confirm. We respect anonymity, GMX is an anon protocol, we also don’t want the identity of a member of a respected DeFi protocol like synthetix to be misappropriated by an imposter seeking to sow dissent between our protocols. If we can’t in some way verify this we may need to delete the posts to avoid allowing an imposter to post.

[horse-shaped-smyl]

1. 1.2% in itself I’m not so concerned with. But is there a proposed time-limit after which it would be renewed by DAO vote? (possibly with re-negotiation)?

2. What would be the exclusivity period? Once the oracles are proven-out Chainlink will want to offer them to everybody.

3. Regarding Pyth, I know nothing of this situation. However, if I’m a competitor (possibly with VC backing) of course I want to establish a foothold in the market by offering my product with huge discounts or free. But at some point everything needs to be paid for and profit generated.

[DaDao]

Why we don’t use the price from UniSwap, 1inch and other DEX? That would be free.

[ArmyPop]

As a 70 year old consumer will little knowledge of how this platform works I must say GMX has been awesome. Being frustrated with the current banking system I was shown GMX. After a year I am very happy with how it has preformed. I leave the changes and upgrades to those with the knowledge. I trust GMX to do what is best for the community. Paying a set percentage for life eventually could eat up profits. I like the concept but feel it should be tiered. Maybe a year end performance bonus? In the end if it is good and healthy it will multiply. If greed starts making decisions, many will leave no matter what it pays. The platform will no longer be for the people or a safe store of personal wealth. That’s why GMX exists, right?