Mårten Blankfors
Mårten Blankfors
> You need to update the construction of the signer file to include the database name. i.e. add `db_path = ":memory:"`. See build_signer_config_tomls Done ✅
> Could also add some tests for in memory rather than using files :P Sure. I added another test now. It feels a bit outside our domain to do this...
For reference, here is the commissioned report ([link](https://github.com/stacksgov/sips/blob/11f12e52d546d781dd0405de605f17df81d7347d/sips/sip-021/MEV-Report.pdf))
Superseded by https://github.com/FuelLabs/fuel-vm/pull/820
> Yeah I do follow that reasoning, but when signing for sBTC, will the type need a `block` field? Am I wrong that it would make more sense to have...
> I think we don't need to include example corpus in the `fuel-vm` repository. The CI action could do all steps to produce corpus. > > @netrome The final goal...
> LGTM!=) I can't approve since I created this PR initially=D Nice, I'll approve in your place=)
> perhaps we want to benchmark the difference in performance :) Yep, running the `db_lookup_times` benches right now.
> > perhaps we want to benchmark the difference in performance :) > > Yep, running the `db_lookup_times` benches right now. I can't see any conclusive difference between the implementations...
> I can't see any conclusive difference between the implementations with the existing benchmarks. There's a high degree of variation between the runs and the benchmarks are mostly measuring retrieval...