proposal: allow add name to column

[lispc]

to methods like instance_column , advice_column to help debugging

[CPerezz]

I think these kind of additions might be also benefitial in the upstream impl (in case they want them).

In any case, this is a quite big change so I think we should first check with @kilic who mainly takes care of rebasing upstream regularly to see how this would affect his work and if it would significantly increase conflicts at rebase/merge time with upstream.

cc @barryWhiteHat as we just talked about this topic.

[CPerezz]

@lispc I started some work on something similar to your proposal. See: zcash#706

[jonathanpwang]

Column<_> implements Copy, so you probably don't want to add a string or even a &str to the struct itself. But you could have the ConstraintSystem keep a separate vector that maps column indices to column names, which the MockProver has access to.

btw ideally do not use dyn closures for passing strings since that adds a slight runtime overhead.

[CPerezz]

Column<_> implements Copy, so you probably don't want to add a string or even a &str to the struct itself. But you could have the ConstraintSystem keep a separate vector that maps column indices to column names, which the MockProver has access to.

btw ideally do not use dyn closures for passing strings since that adds a slight runtime overhead.

If you check the impl, is not the Colum who holds a String but rather the Region struct. As the naming is done at synthesize time.

Also, from all the Assigment trait implementors only MockProver does hold the namigs. Which meanst ghat no overhead happens for the real Prover.