Transcript of the Ampleforth office hours on 16.06.2021
Simon:
Thanks for joining today’s office hours — it was a big week with the first ever cross chain implementation of AMPL as AMPL made it onto BSC helping fulfil the vision outlined a few years back of an independent currency in a multichain world. Pancake Swap quickly followed on with AMPL support including a Syrup pool you can read more about the 2 incentivized options involving AMPL on Pancake Swap here.
Those are exciting steps in the history of Ampleforth which continues to build and develop — AMPL is 2 years old this week after all! But more is being done. You may have seen in the Governance channel discussion around the Tellor proposal. There has been great conversation on the topic and also some questions. The Tellor team has been a good group and has some history with Ampleforth so it made sense to include Nick Fett from Tellor to join today’s office hours session to discuss. Let’s take this opportunity to discuss AMPL, Oracles, Tellor and any questions you may have for Nick!
Question 1:
Any plans to include PancakeSwap in the market oracle ?
Aalavandhan:
Not yet but I think we should soon have it in there. The chainlink oracle currently uses data from aggregators and once those aggregators add pancake swap, we should see it make its way into the price feed.
Question 2:
Will Ampleforth launch on any BSC lending dapps anytime soon?
Brandon:
A big part of how we determine how to prioritize new chains and platforms is how much of an existing ecosystem there is for us to integrate with. Borrow/lend is obviously very high on that list.
Question 3:
What chain does the team anticipate being next? Will bridging to other chain be easier now that it has been done for BSC?
Aalavandhan:
Matic should be next (FYI, nothing is guaranteed). Newer chains EVM which we add through MeterPassport should be plug and play.
Question 4:
Its been 50 days since Forth was released into the wild. Are there any plans that have been laid out ? This is important because the cost basis of the airdropped forth was 35$ on that day . Some of us have to make a decision to sell part of the forth to pay the govt taxes .. or conjure up the money to pay for that from somewhere else based on the direction of forth.
Brandon:
Can’t give investment or tax advice for you, but for plans around FORTH and how we think about governance today I’d recommend checking these out if you haven’t yet:
- https://www.ampleforth.org/governance/
- https://medium.com/ampleforth/the-state-of-discretion-and-governance-forth-ec2f710d2635
That said, these are systems that can be modified through time, as the system and community allow for it. We’ve already seen a couple proposals on how governance could continue to develop.
Question 5:
With the discussion around Tellor integration. I believe one of the concerns was centered around “bad” data either corrupting or not being used by ample a majority of the time. What has the team be discussing or considering for proving this out?
Nick from Tellor:
I can answer what we’ve talked about with the team. We’re still in talks with Ahmed on working out the quality of the data. Part of the issue with Tellor is that it’s not just one implementation or just us running the code. It’s a network of miners, so we have to push out the updates and at a last resort, use the dispute process. We’re actually still working out what “bad” data is. Is 1% off bad data? How about .5% if LINK is also .5% off? It’s sort of tough to say that there’s an exact right answer, since if there was….you wouldn’t need an oracle. But a lot of those questions are ones for the AMPL team and we’re just trying to work with them to implement the oracle methodology the best we can.
Brandon:
The question of “What is the perfect data?” Is actually very tricky. VWAP? TWAP? Which markets? What size sliding windows? Do you use intermediate pairs? It’s complicated.
I view the existing feeds (including the Ampleforth feed) as approximations of what the true latent value is. But I’m not sure that there exists the perfect number. What we ultimately care about is whether the data represents the market or not and how manipulatable it is in the marketplace. We’ve published guidance on what we consider best practices and are happy help people implement this.
Question 6:
Are you reliant on Meter-passport adoption first before AMPL can implement new chains? I.e Meter adoption = AMPL adoption.
Aalavandhan:
For Matic/Arbitrum for example which have their own native bridge we don’t need to use Meter Passport.
Question 7:
What benefits will Tellor bring that the other price feeds don’t?
Nick from Tellor:
I think the big thing Tellor brings that LINK or the AMPL oracle don’t have is some liveness/ censorship resistant guarantees. It’s a different kind of oracle that doesn’t rely on a central group of individuals to run it, so anyone can become a data reporter and you get some crypto-economic guarantees that you can get a price up there without anyone being able to stop it.
Brandon:
As far as the for the Ampleforth protocol, the overarching goals are still to maximize decentralization and have as many data sources as possible. Systemic risk from reliance on any particular ecosystem is something to always keep an eye on.
Question 8:
One concerns with Tellor is a scenario where we are paying for data usage that doesn’t get used half the time, and if a bad actor attempted to run up the price of the tip/bribe, maybe it’s not a big deal now, but when the price of Tellor is higher that could be a problem.
Nick from Tellor:
Sure. It could definitely happen. I think ideally LINK, Tellor, and AMPL would all be the exact same values. I think just the ambiguity in the docs on sources and the specific API differences make up the errors, but you can assume that if there was a perfect API, we’d all have exactly the same number and I’m guessing that in the next year or so as exchange API’s get better and these moving averages become more accurate, we will all be posting the same number. Once that’s the case, it’ll be more expensive to have any more than 1 oracle, and the exact reason you want more than 1 isn’t because you care as much about a .5% difference, but you care that someone could take the oracles down, or someone (or team) could cheat it and profit. That’s why having a Tellor which isn’t a reputation based system like others, adds some security.
Question 9:
When does the team expect Geyser v2, Aave aAMPL, aTrinity? next six months?
Brandon:
Geyser v2 front end development is coming along nicely. The main functionality is in place, but we do want to add a bit more around vault management. Aave’s aAMPL is still in the pipe, but since the timeline isn’t directly in our control right now I don’t want to make any promises.
Trinity collateral in the Aave AMM market is its own independent proposal action. (Thanks for the initial post btw!). This could in theory happen independently, but I’d like to first focus on getting aAMPL on the main platform myself. The AMM market could be a nice secondary step.
Question 10:
While waiting for Aave, can the team try to get listed with Compound in the meantime? What happened to the Compound/Ampl partnership(cAmpl)
Aalavandhan:
Know any COMP whales? More seriously though I think they are the most conservative protocol out there. I’d imagine we’ll be on Aave before Compound.
Brandon:
My understanding is that Compound is (interestingly) restricted by their oracle. They can only get price feeds through their Coinbase oracle, so they’re restricted by assets available there. I’ve heard they’re doing some work to remove this constraint, but nothing concrete yet. Nick, maybe you can help them out.
As far as the Compound system itself, AMPL works there out of the box. No integration required. Of course, it would still need to go through their normal governance processes.
Nick from Tellor:
Yeah Compound has had a pretty hands off approach on their centralized oracle. We’ve been in talks, but now there’s a proposal to use LINK and we’ll see if it passes (I don’t think their core teams really sold on it). If they want to use LINK (and just LINK), it’s kind of a different direction that their community wants to take.
Question 11:
Right now Tellor has a roughly 2.5x greater standard deviation around the AMPL oracle than Chainlink. What would you consider an acceptable deviation?
Nick from Tellor:
So we’ve definitely been making changes and updates as we talk to the team (and to you guys the community) and we plan on continuing to push to our miners what “acceptable” is. But ultimately it’s a decision for you guys and we’ll do our best to match up. We have disputes that we can use to get it perfect (so miners can actually get slashed a lot of money for reporting incorrectly), and we’ll use it as needed to get the price correct.
Question 12:
Are there plans to financially incetivize forth holders for voting to prevent market cornering and centralization ?
Nick from Tellor:
We looked into doing that with Tellor..unfortunately if you pay people to vote, you just get people voting with automated software or more likely to not hold their tokens (or delegate if they can)… doesn’t really fix it. Ultimately a community that’s aligned with the vision and cares is the hard thing to build.
Brandon:
I’m inclined to agree with Nick on this too. It’s definitely something we considered. I’m not a huge fan of rewarding people to vote, though. You might get higher participation, but with lower quality measured intent and likely poorer outcomes.
That said, it’s technically feasible. It would basically entail building a gauge-like system (similar to Curve DAO) that diverts some portion of the yearly inflation to voters. I think there are better uses for this stream, though.
Question 13:
Off-chain signalling should be replaced with node based signalling with delegation . Otherwise we would never have enough Forth holders voting and only the interest of a minority with a louder voice will see their agenda forwarded.
Brandon:
Signaling right now just uses normal FORTH token balance, but this is only temporary until we have a tool that makes delegation easier. This should be coming very soon, I’m guessing within 1–2 weeks.
I agree. I’d like to see delegation as quickly as possible.
Question 14:
It would be great if Tellor could publish a live feed/graph that compares your data to the Chainlink and AMPL oracles on an ongoing basis so we can judge your progress. Should be trivial using the existing APIs.
Nick from Tellor:
We actually have one (not the prices on chain, but the ones the miner is calculating that would be pushed on chain if you request it then). We’ve gotten it really close over the past few weeks, but I’ll pass it on once I find it (our dev krasi manages the monitors).
Aalavandhan:
Not just for AMPL, but I think would be great if someone could build this for even ETH, BTC price. Compare price inputs from Uniswap TWAP, Chainlink, Tellor and Band.
Nick from Tellor, in response to Aalavandhan:
You could, but just remember definitions matter and a successful centralized price is also no accomplishment (I’m sure coinbase’s API works just fine).
What a BTC price is could mean a million different things, and knowing the cost to alter the price (manipulate the exchanges), or the cost to censor the oracle (or the awful idea of just trusting someone)… it’s a really hard comparison to make (we’ve tried in the past).
