Robert Kuykendall

I don't use refs often, so perhaps another user can help me understand the issue here, but here is our test for the innerRef feature if that helps: https://github.com/formsy/formsy-react/blob/49fc63403ea7d9d712a993114873b8abb26c8686/__tests__/Formsy-spec.tsx#L11-L32

Can you branch off and write a failing test, similar to the one shown above?

I don't like to bump NP to above 5 because 6 requires you to use 2-factor on NPM. That said... I do recognise that is a good thing I should...

Thank you for reporting this! All changes from 2.0.0 to 2.0.2: https://github.com/formsy/formsy-react/compare/0c681fd77ce33a49d2d048e44e79f9404af851c5...838cfbd9c36944ae500eea9e54a23d6569d857e9 There's a lot in here, but we can ignore test, type, dist, and build changes, so that narrows...

@hungbang There is currently no solution for this issue, but PRs are welcome as well as failing tests or even just edits to `API.md` with the desired feature spec. I...

A pull request would be very welcome for this feature.

If we renamed `isChanged` to `isPristine` and added it to `API.md` does that solve this feature? A test would be nice too, if it became a publically documented function.

So in summary, onChange should only fire if the values differ. This seems like something we could probably write a test for and then implement internally without breaking anything or...

That's an interesting point. Can you just surround your `updateInputsWithError` with a try-catch? The downside to warnings is they can't be caught and handled.

That is a very good point. I think that would be an easy refactor to update all found fields and their errors for the rest. I'll work on it.