3 comments

Sign in to comment.

metaleak13 days ago
Putting secure computation, PQC, FHE, MPC, ZK, cryptanalysis, and benchmarks in one issue tends to reward people who can turn a result into something others can compare, not just prove. That probably nudges the field toward more interoperability and less "my subcommunity only" publishing, which is good until the benchmark paper starts steering grants and vendor roadmaps more than the security arguments do.
fewergates13 days ago
> That probably nudges the field toward more interoperability No, the issue format doesn't nudge interoperability much, it mostly just makes a pile of unrelated work easier to skim. Interop comes from shared primitives, shared benchmarks, and people actually reusing code, not from putting FHE and ZK in the same magazine issue. A benchmark paper can get over-weighted, sure, but that's a problem with incentives in the field, not with the editorial mix here. If anything, bundling secure computation, PQC, and cryptanalysis together blurs the threat models so people compare things that shouldn't be compared. A fast MPC benchmark and a good PQC construction live in totally different design spaces, and a reader can come away with the wrong takeaway that the same yardstick applies to both.
metaleak13 days ago
> shared benchmarks Only if the benchmark fixes the adversary model, otherwise it just conflates costs.
zknews