You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Question
We recently noticed a nonce increment issue on pre_disptach. It seems like nonce does not increment during pre_dispatch but rather increments after nonce. This is bit of deviation from Substrate where nonce increments during the pre_dispatch.
I think I understand that EVM does this automatically but will there be an issue if the nonce increment is moved to pre_dispatch instead ?
In Subspace, we have an EVM chain as an enshrined rollup. We do pre_dispatch before including the transactions into the Subspace chain block and once the block is imported, we derive the eth block from the Subspace block.
To ensure only valid transactions are included, we do pre_dispatch. Unfortunately, if there are any transactions with future nonces, all those fail at pre_dispatch.
Are there any concerns or potential solutions to make the pre_dispatch behave like Substrate instead ? This would not be breaking any existing chains since the pre_dispatch and execution are done at the same time.
The text was updated successfully, but these errors were encountered:
Question
We recently noticed a nonce increment issue on pre_disptach. It seems like nonce does not increment during pre_dispatch but rather increments after nonce. This is bit of deviation from Substrate where nonce increments during the pre_dispatch.
I think I understand that EVM does this automatically but will there be an issue if the nonce increment is moved to pre_dispatch instead ?
In Subspace, we have an EVM chain as an enshrined rollup. We do pre_dispatch before including the transactions into the Subspace chain block and once the block is imported, we derive the eth block from the Subspace block.
To ensure only valid transactions are included, we do pre_dispatch. Unfortunately, if there are any transactions with future nonces, all those fail at pre_dispatch.
Are there any concerns or potential solutions to make the pre_dispatch behave like Substrate instead ? This would not be breaking any existing chains since the pre_dispatch and execution are done at the same time.
The text was updated successfully, but these errors were encountered: