is there a breakdown/comparison of Elpi vs. TC and CS ("less ad-hoc" paper) specifically for Prolog style backtracking proof search? As in, performance, lines of code, etc.? Or is this what the Rosetta stone might have in the future?
No really. Davide Fissore (@Davide F) is writing his Master dissertation, for now the only (very incomplete) bench is here: https://coq-workshop.gitlab.io/2023/abstracts/coq2023_TC-elpi.pdf
He already has some nicer graphs, maybe he can just post here a screenshot.
For a less "crude" comparison I think we need to wait for him to complete this report.
Last updated: Oct 13 2024 at 01:02 UTC