@witty-crayon-22786 is this materially different from your view? it moves the part that the engine has to interpret specially from checking the subject’s type to (in this version) introducing a separate FFI object on the rust side (i’m assuming — @hundreds-breakfast-49010 @hundreds-father-404 let me know if that’s correct). does that address the syntax concern here, or is there another direction you’re thinking of? i’d like to note to stu that this also shows greater buy-in from toolchain regarding generic retry as a concept and possibly worth reopening the syntax discussion to implement it. stu left the office today not feeling well so i wouldn’t expect him to be able to reply today