refactor: make solution extrinsic mortal #728
Conversation
| nonce as u64, | ||
| Default::default(), | ||
| )?; | ||
| let signed_phase_len = client |
There was a problem hiding this comment.
NOTE: this is static applies the extrinsic to live for "signed_phase_len" and if for example the last block in the sign phase it will "live" for additional blocks.
We could change that as well but that is a slightly bigger change.
There was a problem hiding this comment.
Yes, I think this approximation is OK given that we also have the checks on the EPM side.
otoh, maybe it would be easier to know exactly the boundaries of the signed phase in on terms of block number if the Phase::Signed in EPM wrapped the block number when it started? That way the miner could fetch the current EPM phase and know mortal(signed_phase_started, signed_phase_len). is this worth the change?
There was a problem hiding this comment.
Yeah, indeed it requires some changes because each block is executed in its own async task and needs synchronization.
Lemme create another issue for it :)
There was a problem hiding this comment.
Thanks! I think for now this and the EPM-side changes are enough to tackle the original issue.
| nonce as u64, | ||
| Default::default(), | ||
| )?; | ||
| let signed_phase_len = client |
There was a problem hiding this comment.
Yes, I think this approximation is OK given that we also have the checks on the EPM side.
otoh, maybe it would be easier to know exactly the boundaries of the signed phase in on terms of block number if the Phase::Signed in EPM wrapped the block number when it started? That way the miner could fetch the current EPM phase and know mortal(signed_phase_started, signed_phase_len). is this worth the change?
Close #727