Given the high gas price environment on L1

Currently, ETH L1 is at full capacity and gas fees are “elevated”. There are currently many L2 solutions coming for scaling, etc.

For discussion purposes, is the objective to keep the core of MKR DSS on L1? or to have zksync for DAI transfers? xdai? or just let the market decide?

Personally, I am just starting to learn the basics of zksync and roll-ups in general. I am curious to get comments from the community on zksync vs xdai. etc… and more importantly how the DAO can use L2 to scale in general.

Further, how the arch. will evolve given a future w/ sharding?

I would also assume the SourceCred team would want to start using this as well given the elevated fees.

Vitalik’s comments on rollups.


My understanding is that for trading assert, market time is very important therefore Dai has to be on L1 but exchange can go on L2 with it.
For normal use Dai can go on L2 that is probably the xdai market.

For the dss the main use is deposit assert withdraw assert, I am not sure how much you can put on L2 as assert are on L1. I am not even sure if there is a lot of traffic. Probably relatively low.

Governance can go on L2 tho.

But that is just a high level view with little inside.


Part of the reason I am asking is that from my understanding and research (which is very elementary now), some of the EVM processing can now be pushed to L2. Curious how that impacts DSS and governance… and how the community can leverage this.

I think we can start thinking about in a Core Unit of L2 solutions, protocols like Uniswap and Synthethix are joining to this solutions (zkRollups, Optimism …) cuz we all know that ETH 2.0 will come in some years, so not having a team focused on take the Maker Protocol to some L2 solutions is just not moving with the flow and being stucked in the ETH mainnet, also leaving asside tons of users with low capital.

I always says, one of the principles of MakerDAO and Dai is Serving The Underserved, and we are not giving solutions to them.


I tend to agree. While folks can use rollups outside of maker (and that can happen today), increasing participation on voting is important. Personally speaking, I would think voting on L2 and then allowing anyone to bump the L1 chain with the outcome would be a great way to encourage voting.


Right now even voting with snapshot, at least on just Governance Polls and doing on-chain votes on Executives ones can solve the problem in short-term, but yeah, voting on L2 and start with oracles on L2 is a big step for Maker in order to prepare for the future before ETH 2.0.

1 Like

We’ve been able to tame the gas fees somewhat using, which batches the transactions in a gas-optimized way. However, the latest on-chain payout last week cost ~$275 worth of ETH in gas, the most we’ve spent to date. We sent out ~$25k worth of DAI, so that’s ~1.1% of the total.

A few contributors have asked about xDAI in particular. I’ve been hesitant to explore that, mainly because it’s relatively new, and I’m wary of introducing more friction to the process of participating. That said, I’ve been using xDAI for a few months, and it seems to be maturing quickly. Also, 1Hive, which runs on xDAI, has been distributing its token $HNY via SourceCred for a few months now without complaints. Maker is already affiliated with it…might not be that big an issue. Would be curious to hear from any SourceCred participants, do you have opinions on being paid on an L2? Objections? Preferences? Other L2s you think are worth exploring?


I know already supports DAI… No objections (personally speaking). I think it is great to help “nudge” the entire ETH community to L2 for smaller transactions.

We are doing a lot of cool stuff on xdai at 1hive. Building out bridges to matic, and other L2s, actively trying to unite L2 into a space that feels and acts more unified. Xdai itself has come a long way recently with the switch to dpos, and the L2 space is maturing rapidly.

I would love to get my SourceCred payments on xdai, as it now often costs almost $100 to do anything on mainnet, and xdai is growing and building more defi infrastructure all the time. We would love to have Maker cdps on xdai as well. Although maybe that is outside the scope of this discussion.


me too. I have proposed this several times.

I think this would be amazing! But probably it would be necessary to – automatically – move to Ethereum mainnet the collaterals of the Vaults, very frequently (batching the transactions would lower the fees).

It seems complicated and would definitely require some serious smart–contract competences. Currently MakerDAO does not have the bandwidth to do this. We need external people starting a dedicated Core Unit and do this:

See this thread.

@befitsandpiper please spread the word.


We’ve been pushing the xDai p2p market on Venezuela (where i live), but i have some question from community:

  • What do we say to people saying xDai is more centralized than the mainnet where they have their DAI.
  • Also the bridge between L1 and xDai incurs in gas fees as well, that itself apart some people from using it (until they see how low the fees are xD), wouldn’t that be a problem too in order to distribute the Sourcred payouts?
1 Like

xDai is technically more centralized than mainnet, as it runs on dPoS instead of PoW or PoS. But this is part of the blockchain trilemma. There is no true L2 solution that is available at the moment, which would be capable of overcoming the trilemma.

Moving to xdai chain is essentially a vote that dPoS is sufficiently decentralized for smaller players and communities to use in exchange for nearly free transactions. xdai is also looking for new ways to improve security and decentralization, including optimistic bridging.

The way that you could reduce SourceCred gas fees for distributions is by moving all funds on distribution day over to xdai in a single transaction before distributing to individual wallets. This would require only a single mainnet transaction, rather than n transactions where n is the number of wallets being sent funds.

1 Like

It is all about risk management. Deploying on xDai is OK. I think. But i would consider it as an experiment. A low cap for xDai on WETH and a modified Maker naming could be considered.

Agree that would help out SourceCred–but personally I have no interest in xDAI. I would immediately go back to Ethereum Layer 1 as I trust it 100x more and it has a lot more composability for my needs. And to leave xDAI it would cost me ETH, and it will incur an unnecessary Taxable Event (xDAI swap to DAI is consider acquiring Property A for Property B).

However, if some Community Members prefer xDAI than they should get SourceCred via xDAI. Not everyone has the same User needs as I do–so perhaps folks should choose what they prefer.

Hopefully the xDAI experience will blossom and I’ll hop over and enjoy it as much as others do.

xDai and Dai is 1 for 1 so it’s not really a taxable event cos there’s no gains…

Maybe people could choose how they get distributed yeah. Personally I find the fees on mainnet much too high for my needs and xdai security sufficient. But yeah, that should be a choice each user can make based on their own needs.

I think for many small players fees on mainnet can eat 1% or more of their stack from gas alone in a single transaction (basic swaps on uniswap can easily be $100 these days, and more complex contracts or multi-step processes can be still more). That problem doesn’t exist on xdai because transactions are essentially free.

I can see why people with $1 M + would want to stay on mainnet. Even $100k+ I can see reasons why. But it’s starting to become really difficult to justify doing anything on mainnet with less than 6 figures. And gas fees will only get worse from here for the foreseeable future.

Matic is also an option. There is an xdai/matic bridge being built now. Many of these L2 solutions and sidechains will likely be bridged together in the coming months.


Personally DAI on ETH has more utility for me.

It’s the dilemma where if you send on xDAI, people will need to spend more gas to unwrap onto L1.

If you send on L1, some people might need to wrap onto an L2.

1 Like

I second this.
I’d prefer much more to have something built on a generic solution based on rollups (ex. though, which are still secured by Ethereum rather than another sidechain. Nothing against them, but it just feels like a more clean solution to use rollups.

EDIT: Just saw zkSync is working on a generic solution, not “only” to send tokens around


Hello @ianaz and welcome to the Maker Community!

I agree–something tells me that some of these “sidechains” are going to become their “own major chain” once they get some traction. And if Major Exchanges, or Fintech Apps can allow Users to deposit Fiat directly to the “sidechain”–then there won’t be a need for the average human to use base layer Ethereum (until we get ETH 2.0).

So, yes these are some my concerns with regards to sidechains.

Also, we should look to expanding adoption of DAI within other Layer 1s. I was just reading that TerraLuna’s U.S. Terra (UST) stablecoin is accepting Solana, Polkadot, and StakedETH as collateral. All can be used as collateral to mint UST. The competition is getting thicker in the stablecoin ecosystem.


This topic was automatically closed 91 days after the last reply. New replies are no longer allowed.