Mainnet forking

[fracek]

As I mentioned in the community call last week, it would be amazing to have mainnet forking. The use case it's to test how our dapp interacts with existing dapps deployed to the network. I know that now it's not possible to implement this, so this is more of a question to know what is needed to make this happen?

[FabijanC]

Thanks once again for the proposal. I'm sharing this issue to more places.

[0xSidius]

Any updates on this important feature? @FabijanC

[FabijanC]

Any updates on this important feature? @FabijanC

I agree that it is important. It is in plans to be done, but not in the very near future.

[ivpavici]

On hold until: "we have an independent Cairo VM + a StarkNet runner with a more lenient API (I'm guessing that this will be available in O(3 months))."

[0xSidius]

On hold until: "we have an independent Cairo VM + a StarkNet runner with a more lenient API (I'm guessing that this will be available in O(3 months))."

do you plan to use the Cleopatra VM (rust) ?

[ivpavici]

Hi @0xSidius ! Could you give more insight how using Cleopatra on our side would help with forking?

[0xSidius]

Hi @0xSidius ! Could you give more insight how using Cleopatra on our side would help with forking?

it wouldn't w.r.t. to forking, but it would significantly improve DevEx, slashing down the current main Starknet bottleneck (performance)

Did you take a hard look at the developer experience on Starknet today? Even a modest project, using the fastest toolchains (Protostar) with Devnet will see hours-long test runs

[ivpavici]

Ok, but this is for another issue... Btw. do you have any estimate how much would using Cleopatra potentially speed up the devnet?

[0xSidius]

Ok, but this is for another issue... Btw. do you have any estimate how much would using Cleopatra potentially speed up the devnet?

if our Ape/Nile vs Protostar comparison is any guide, we are looking at an average of anywhere in between of 1.5x to over 3x