You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The problem appears when both the constraint system and witness are obtained by deserializing data.
Maybe add a new package which imports all hints in gadgets (backend/hint/all etc.) which has only a single side-effect of importing all hints from gadgets and registering. This package should then be imported by the compiler and provers.
The text was updated successfully, but these errors were encountered:
not a big fan of the registry pattern here (well the one for the #245 too :) ).
I think the R1CS / SparseR1CS when it calls solve, should be able to check which hint it was "compiled" with (at least a string name with package) and output an error message to say that said hint is not injected. But I'ld rather have the user do "r1cs.Solve(witness, mygadget.MyHint...)" than to import a package for its sideeffect.
Let's sit on that one a bit, it seems there are couple of refactoring change on the horizon that needs careful thinking 👍
The problem appears when both the constraint system and witness are obtained by deserializing data.
Maybe add a new package which imports all hints in gadgets (
backend/hint/all
etc.) which has only a single side-effect of importing all hints from gadgets and registering. This package should then be imported by the compiler and provers.The text was updated successfully, but these errors were encountered: