R. C. Howell

Why draft?

> So, this translates to deferring failure until evaluation. The specification says most commonly you want to error, but _can_ defer. This is what [Yingtao's PR for planner modes](https://github.com/partiql/partiql-lang-kotlin/pull/1385) handles....

I still don't quite understand the purpose here other than conformance tests being wrong. I believe Yingtao's planner work and the compiler implementation are correct. **HOWEVER** I think perhaps we...

Edit: A Converter is not a Visitor; A Converter is generated for each Sum type, so it does not serve the same purpose as Visitor

Are you interested in keep all three "visitor" types (Visitor, VisitorFold, Converter)? As a new person looking in, I see that all three of these are covered by Visitor

I see how Converter is an interface, so you have to implement them all. Was that intentional to enforce? If I were writing a converter for say "uppercase all identifiers",...

Understood. I am new to the team and was taking a look at this package and saw some potential to consolidate classes and logic. It's not prioritized now.

> The style of visitor you list there requires that all child nodes are contained within a children collection, which PIG does not generate. Instead, each child node is a...

I'm curious how this is different than using accept? I believe the intention of an `accept` method is to invoke the appropriate visitor method when then concrete type in unknown....

If visitors have parameterized return types like discussed here -- https://github.com/partiql/partiql-ir-generator/issues/123 then you include the parameterized return type and context objects in the accept. The current state of visitors/transformers/etc. is...