Victor Barua

This seems pretty interesting. Are there any more details for this project or is it still in early development?

> In other words, if the field is not set, it will take on its default value. The consumer cannot distinguish between the two cases. That's exactly the issue. We...

The generated Java code doesn't have a method to check nullability either.

PR to address this here using -1 as a sentinel value: https://github.com/substrait-io/substrait/pull/627 One question that I think is worth asking. If we were able to make a breaking change, would...

> Correspondingly, consumers tend to ignore this field. So substrait-java doesn't ignore the field, as I discovered during our version update internally in which plans using `max:tstz` in `functions_arithmetic.yaml` no...

@joellubi > I believe you were focusing on the right-hand side I'm definitely rhs focused as well 😅 For your example starting from SQL, consider something like: ```sql SELECT avg()...

> perhaps an option in common I think adding an option for this will result in more divergence in how systems handle it. There is also prior art in how...

I think it makes sense to have this functionality in substrait-rs and then have it be used in DataFusion. That would also allowed other Rust-based systems to leverage it. IMO...

> it seems like [substrait-validator](https://github.com/substrait-io/substrait-validator/) is using antlr-rust, which is both unofficial and unmaintained at this point, but maybe it does the job... That both seems reasonable in an "If...

The changes in https://github.com/apache/datafusion/pull/13127 also addressed this.