Cody Gunton
Cody Gunton
Suspected major culprits at the moment: transcript is all in a header; zeromorph is all in a header; new Univariate template parameter; use of MANY different tuple types in relations.
This is a forever issue that we return to periodcally 😄
I spent a little bit of time investigating this in February 2025. _Any_ change (even to a comment) in bb.main would about 50s to compile on personal dev client machine....
@iakovenkos would you say this is handled in an indirect way by some of the work you've done? if so will you tag and close as appropriate?
We need the non-native ops of the ECCVM recursive verifier to execute quickly in the simulator. This might require additional work.
Simulator is deprioritized.
You note also that there are some sha and blake tests that are also commented out.
Still an issue https://github.com/AztecProtocol/aztec-packages/blob/5cde0a11e77de304842090d33d728af9e2a085f2/barretenberg/cpp/src/barretenberg/stdlib/primitives/plookup/plookup.test.cpp#L26
Simulator is deprioritized.
Maybe none of this needs to be done except for the merge part, which needs to be handled through folding because we don't have a proof of correctness otherwise.