barretenberg

barretenberg copied to clipboard

A mechanism that treats the whole circuit before finalization (in case of ultra) as a graph with additional information and looks at the following properties: 1) How many independent subgraphs...

Rumata888

The original version of ClientIvc automatically appended recursive verifiers to each circuit. This was simple but did not allow the proof/vkey witnesses used in noir to be properly connected to...

ledwards2225

Figure out the concatenation trick with Shplemini

maramihali

Adds dev complexity but this is standard.

codygunton

added a hack to sumcheck.cpp that should be undone

ludamad

The work [here](https://github.com/AztecProtocol/aztec-packages/pull/8503) is more complicated because almost every use of `finalize_circuit` does not add the extra gates. If we keep these extra gates then we should change the finalization...

codygunton

Processing RAM/ROM updates range constraints during finalization, so computation might be off

Rumata888

Fix KZG for identity group element

1

Currently gemini breaks for a polynomial, where the second half of coefficients is zero.

Rumata888

TODO left by Zac in a few places regarding copying.

lucasxia01

currently get_row does unnecessary copying, could have a row reference class that just bundles a ProverPolynomials and row index (can take a lambda x => x.q_arith etc then apply func(prover_polys)[index])

ludamad