Optimism support


layer 2 networks are excellent soil for the mass adoption of Ethereum. the low fees are suitable for new users who don’t want to be bothered with the technicalities and the ethos of decentralization. yet L2 users will still encounter multiple pitfalls. a huge one is that they need ETH to initiate transactions, while many of them don’t want any portfolio exposure to ETH or handle an additional asset. this is where the Gas Station Network enters the stage: your infrastructure is a major missing piece in realizing the ideal L2 user experience.

Optimism is a forerunner of optimistic rollups and, based on my inquiry, there is currently no effort to tackle gasless meta-transactions on the network.


I want to kick off a discussion between GSN and Optimism developers about the viability of deploying GSN on Optimism. some initial question that come to mind:

  • what changes are needed to the GSN contract infrastructure to work on Optimism?
  • what changes are required in GSN’s off-chain architecture?
  • are current relay server operators open to the idea of expanding to a L2 network?
  • could the relay server role be merged with Optimism’s sequencer role, as to simplify Optimism’s ecosystem design and make it more profitable to run a sequencer?


I’m posting this thread to both GSN’s and Optimism’s Discord (which I can’t link due to hyperlink limits for new users but you can find the link to both on the respective project’s website).

1 Like

Hello there! I’m with Optimism. I thought I’d give a short answer that might help:

We strive to be EVM-equivalent, and so the number of differences should be minimal. We have a list here:

Another thing not mentionned on that page is the fee scheme that has a “L2 fee” paid via the normal gas mechanism + an extra “L1 fee” that depends on the L1 gas cost and the amount of calldata used by the L2 transaction: Ethereum Differences | Optimistic Ethereum Docs

Finally, note that we forked Geth at some point before the Berlin hardfork and so we don’t support the most recent EIPs. In particular, there are no access lists and EIP-1559.

I can try to have a look at the architecture page a bit later. It would be better if someone that has a better in-depth knowledge of the current stack could look at it, but we’re all (understandably) super busy at the moment. That being said, if you specific questions, that’s much more easy for me to relay and get answers to.

1 Like