Honey Migration to Ethereum (signalling/suggestion)

[willjgriff]

Honey Migration to Ethereum (signalling/suggestion)

Proposal Information

Proposal description:
Tentative proposal for gathering interest in migrating to a secure L2 solution, specifically Arbitrum in at least the short term. The general idea will be to migrate Honey to Ethereum, requiring manual conversion by all Honey holders through a UI we will build, and 1Hive Governance to Arbitrum. This structure will allow us to move 1Hive’s governance and general operations to different L2’s without making users migrate to a new Honey token again in the future, should we want to.

See this hackmd for more details, feel free to leave comments, and join the Arbitrum chat in Discord for discussion.

Proposal Rationale
The future of xDai, where Honey is native, is unknown, therefore the future of Honey on xDai is arguably insecure. There is a chance xDai will upgrade to a Layer 2 at some point but details surrounding this are yet to be made. In the mean time Layer 2 solutions are becoming available, the most promising of which seems to be Arbitrum.

Celeste, which is necessary for Gardens, partly inherits its security from the chain where it’s token is issued. Celeste’s security is also influenced by it’s number of keepers. If we move to a Layer 2 solution it will increase security and should increase the availability of keepers in Celeste, ultimately increasing interest and use of Gardens and other 1Hive applications.

Note sidechains such as xDai are likely to be cheaper to run on than Layer 2’s although I expect the difference will seem negligible relative to regularly hiked gas prices on Ethereum. And even in this case, I believe the increased security and likely wider ecosystem interest is worth the cost.

Expected duration or delivery date (if applicable):
2 months.

Team Information (For Funding Proposals)

@willjgriff

Plus whoever is interested in:

• Building trusted minting services, should we choose this approach for token migration.
• Building a UI for converting the current Honey token into the new HoneyV2 token.
• Adapting the current 1Hive apps to support the Ethereum/Arbitrum deployment.

Anyone interested please join the Arbitrum chat in Discord.

Skills and previous experience in related or similar work:
@willjgriff - Solidity development for 1Hive.

Funding Information (For Funding Proposals)

No funding as of yet but to execute we will request some.

[FriedRengi]

Absolutely in favor of this idea. I am eagerly awaiting L2 deployments specifically because of the security guarantees.

I will add my comments on the HackMD.

[Sina]

@willjgriff

Celeste, which is necessary for Gardens, partly inherits its security from the chain where it’s token is issued. Celeste’s security is also influenced by it’s number of keepers. If we move to a Layer 2 solution it will increase security and should increase the availability of keepers in Celeste, ultimately increasing interest and use of Gardens and other 1Hive applications.

Will, hello there. It will boost transaction costs by converting honey to Eth. I believe it will not be profitable for those with limited resources. The majority of crypto are on the lookout for a way to lower transaction fees. I believe that the small capital that makes up such a major part of this group may consider leaving! Thank you for taking the time to write a post.

[Jsh976]

So interesting… I also heard about migrating to fantom or …
It will extend the ecosystem for sure but will be more competetive for us…

[sacha]

To clarify, under this proposal it won’t cost you any more to interact with the 1Hive DAO (or to convert between HNY and ETH). This is about moving to a more secure home (Arbitrum) in a way that gives us the flexibility we need to move again with minimal disruption (should we ever need to).

@willjgriff I think it’s perhaps a good idea to make this clear in the proposal information itself.

[project_uwb]

Will, thanks for the writing up the proposal. I have been having some thoughts about how our future on xDAI looks like and what it would have been if we were on Polygon already. I am not technical enough to understand the differences (in detail) about rolls ups vs side chains and the compromises we have made to be on xDAI.
I am in favour of this proposal as I can see the network affects of a sidechain that has the users and the exposure it gives projects specifically from being on Polygon over the last few weeks, I can see a flurry of activity and innovation to make DeFI and other applications accessible to everyone instead of a select few due to the costs associated with smart contract interactions on ETH main net.

@sacha

To clarify, under this proposal it won’t cost you any more to interact with the 1Hive DAO (or to convert between HNY and ETH)

My concerns though are aligned with Sina in that, xDAI allows me (personally as a smol timer i.e limited capital) to experiment and participate in all things 1Hive such as the ease of being able to swap between tokens on Honeswap with negligible fee, interactions such as staking, providing LP and being able to participate in any activity with negligible costs.
Would this move then lead to a decline in some of those activities as I assume a % of the user group may be priced out due to the tx costs involved which could be significantly higher when compared to xDAI tx costs? It would be nice to see if we could lay down the benefits and then also look at the increase in costs to transact and have a robust discussion on this to keep the community informed if we do take this option to move!

Also read up on the HackMD notes and have some questions:

1. So we are going to have HNY on main net ETH and a copy of the token on L2 which will attempt to mimic the supply on L1? Have I got this right? So why did we move HNY to xDAI in the first place was it not the cost of tx on mainnet? Are we now expecting in the mid to long term for the gas prices to significantly lower? Is this why we propose to move back to L1?
2. I am not tech enough but having a copy of HNY on L2 and trying to replicate it across on a regular basis back to L1 seems pretty complicated - does this create new attack vectors? i.e being able to mint by exploiting this process and rekting HNY?
3. What if some users would like to continue on with xDAI and refuse to bridge over? Does that mean there will be fork of 1Hive across two different networks? What does that look like? Is that also an attack vector?

Would be nice if this could be opened up for discussion on like church call setup, so the tech experts can talk it out and the community can listen in ask questions to learn more? @CurlyBracketEffect as always mate… would be a nice topic for DOAn to Earth!

[Stonky]

Absolutely in favor or this, both for the potential and interest surrounding Arbitrum and for the stagnant situation xDai is in. Thanks will!

[stillnoob]

Thanks for clarifying this. I also thought the movement of 1hive projects a few weeks ago as many rollups are planned to be developed in the Ethereum Mainnet to boost transactions in Ethereum Mainnet. Would be good for 1hive community if we do it earlier.

[gabi]

Will, thanks for starting the conversation. I support the migration

Happy to help with point number 3: Adapting the current 1Hive apps to support the Ethereum/Arbitrum deployment.

[Blazingthirdeye]

@willjgriff

If we move to a Layer 2 solution it will increase security and should increase the availability of keepers in Celeste, ultimately increasing interest and use of Gardens and other 1Hive applications.

How would the availability of keepers be increased by moving to a L2 solution?

I always thought xDai was an L2 but I now see a sidechain is not an L2. I hear that security is relatively weak on xDai and I think it is important that we have sound and safe infrastructure or the community could be torn apart by a vicious exploit. I thought 1hive just moved to xDai and the thought of another move is a little painful. What happens when another better chain comes along? I am sure the community does not want to always be on the go, nomadic style. Yet in the post it already acknowledges that this move is in part to make future moves easier. Wow that would be the third migration. All this planning assumes Ethereum will be the dominant chain indefinitely and 1hive will not want to depart Ethereum Mainnet even if it jumps from sidechain to various L2s.

The workload of migrating is significant. I trust the developers have 1hive’s best interests in mind. I am definitely not qualified to make sure statements about the situation. I wasn’t even aware of xDai’s stagnation until recently. I am just thinking out loud with this posting. As a non-developer I am not sure of how well ‘staffed’ 1hive is with developers. I would like to see a balance of nomadically seeking a better chain solution with continuing to ship great products like the latest Tulip farms and Agave plus Gardens and the usage of Celeste. If we devote all the development to moving the great products will slowdown or stop.

I am very unsure and will read the hackMD for more details. I tend to trust the devs and the senior members of the community but I am also a small fish and want to keep usage costs low like the others have commented.

Thanks for the great details posted here.

[Sina]

How is the transaction’s arbitrum fee calculated?

[Eth_Man]

I consider this proposal (migrating 1Hive infra to true L2 with L1 security etc.) to be absolutely priority #1.

I have been following discussions on Maker forums as they follow this technology closely.

My questions. Are we sure Arbitrum is the best solution and do we commit to this by making the change. My impressions based on my understanding of the tech is that we would finally be homed on L1 and we then will be able to live on any L2.

Any idea of what fees for transactions are going to look like relative to main net? We have a lot of infrastructure on-chain and I wonder how much it will cost a user to interact with celeste or the Honeypot. Is there any possibility to still use pieces of side chains for basic voting and infrastructure using the balances on L1?

Thank you for leading this btw. I believe that it is critical that 1Hive with its technology get to one L2 as soon as possible. The longer we delay the less chance what we have to offer will be useful there.
In conjunction with above proposal is there a list of things that we can bring along with us that may derive income. Are we planning on Honeyswap, Agave, farming, etc. on Arbitrum?

aCOMB anyone?

[eenti]

I’m not the smartest in this room when it comes to this topic but I’ll try to do my best.

The intention is to be safeguarded by the benefits and security that ethereum represents, I think the main one is its decentralization; but then, just as you mentioned, fees would make it pretty expensive to interact for some users.

And that’s what the L2 (Arbitrum) is for. Arbitrum is an Optimistic rollup (i understand these are the only kinds of rollups that support contract execution), so we would work with the benefits of being on this “sidechain” (such as lower fees) that is secured by ethereum basically.

Then, answering to your points:

1. The issuance is moving to L1 precisely because that can help in case we need to move again to another L2, we would have to just bridge the token instead of deploying a new issuance etc.

2. I’m honestly not sure about this lol but I guess an issue is the logistics behind the synchronization of both.

3. Don’t take my word, but I’m guessing HNY on xDai will just behave as the bridged HNY on polygon we have. We’re just moving our governance basically.

[willjgriff]

@Eth_Man

Are we sure Arbitrum is the best solution

AFAIK the 2 layer 2 solutions realistically available soon are Optimism and Arbitrum.

Optimism requires using a modified version of the solidity compiler and they suggest modifying at least constructors to become separate init() functions due to some security risks. The specialised Solidity compiler also increases bytecode size by ~25% and maintains the 24kb limit for contracts. Arbitrum doesn’t use a specialised Solidity compiler, so there is no increase in byte code size and in fact has removed the bytecode size limit entirely (or at least increased it significantly, as observed from my tests), which allows us to max out the Solidity optimiser making contracts as cheap as possible for users to execute transactions on.

Optimism uses a basic token bridging process which would require some custom bridging contracts to reliably fit the needs of 1Hive (specifically that the L2 token total supply should match the L1 token total supply) and it’s unknown if we will be able to integrate our bridge contracts with the default bridging UI. Arbitrum’s bridge contracts are more flexible and we should be able to use the Arbitrum bridge UI.

There’s a few other differences in their general operation, like optimism’s gas per transaction is limited by the L1 limit where as Arbitrum doesn’t have this limit, see more in this tweet storm: Thread by @krzKaczor on Thread Reader App – Thread Reader App

@Eth_Man

Any idea of what fees for transactions are going to look like relative to main net?

Honestly, I don’t know, although I’m sure they’ll be more than xDai. As I’ve said though, we can optimize our contracts to the max possible making gas costs for transactions cheaper than they currently are on xDai, even if the gas price is ultimately higher.

@Eth_Man

Is there any possibility to still use pieces of side chains for basic voting and infrastructure using the balances on L1?

To use sidechains as well as L2 solutions would require more investigation. Ultimately their would be more bridging delays and of course whatever’s on the sidechains would have limited security.

@Eth_Man

aCOMB anyone?

We should do this ASAP. I suggest you push for it in the Tulip Discord.

[project_uwb]

Hey @willjgriff after learning some more about Artbitrum (roll ups vs side chains) and how we are pushing to be on Arbitrum, just wondering what the future of the xDAI related 1Hive ecosystem will be? I mean there is a host of other projects that have found support through 1Hive, Honeyswap, HNY and the whole infrastructure that 1Hive as a DAO provides - would be interesting to know what we leave behind and if the plan of our migration should also consider these interface projects/DAOs?

[Vyvy-vi]

I’m wondering, what ways can be used to better communicate the benefits and implications of moving to people that are not completely acquainted with how things work or are relatively new to 1Hive?

[Blackbus]

I think the value of xdai is great for the 1hive community and so any plan to move entirely away from it is not going to be a win-win. Is there a place in this plan to have an easily crossable bridge to both keep the xdai features functioning whilst also offering the L2 main-net options?

I (and I don’t think I am alone) use honeyswap to buy smaller amounts of low cap projects and don’t really want to convert them all over to L2 individually, losing larger fees if I ever choose to swap them. That said I can see the value of having less roadblocks for traders to come over to honeyswap and the 1 hive ecosystem by being on the main-net.

[stillnoob]

The projects from 1hive will still be available on xDai. Only the governance infrastructure will be moved entirely into Ethereum Mainnet. I think it will be good if we expand the projects from 1hive into the Ethereum Mainnet as the Ethereum Mainnet is having far more serious markets and projects than the xDai projects.

[hongnian]

很棒的提议，迁移是最正确的选择，我很期待

[KZFlyer]

Hopefully this can help. Transaction costs on Arbitrum are definitely more than xDai, but manageable compared to ETH Mainnet.
For some gas pricing context - I recently:
Swapped tokens on swapr for 0.00049330717375 ETH ($1.56)
Added liquidity on Balancer for 0.00069391374448 ETH ($2.20)
According to etherscan right now its around $50 to do either of those on Uniswap.