Mark Elliot

Nevertheless: - all of those are examples of polyglot namespace support - we can provide other, direct discouragement of within-scope same names without making it a hard fence - your...

Maybe I’m misunderstanding your analysis, but if the types that conflict typically span different definitions in different repos, how does the generator flag help?

I think we're more likely (but still uncertain) to end up with globally unique type names when they're namespaced than if we remove the namespace entirely. For your example, it...

Nothing about what I'm describing is Java centric; I'm interested in reasonably unique names across Conjure definitions. In a world where we remove packages from the IR, we lose the...

I think we should be striving for the ideal that types and services defined in Conjure are by and large unambiguous, globally. To do that, types and services defined in...

On the Java end: I think we can revisit the unions-as-objects-with-optionals PR I put up a little while back, it didn't need the type field and generated code had fewer...

All unions encode the value in a specific field name, and the field name can be used to identify the type of its value (defitionally). One might model union types...

@tom-s-powell I think we'd need to do some work to ensure that all deserializers across all the languages get to the point where `type` is ignored and then do a...

Pretty sure we intended WikiCase when this was written, but weren't pedantic enough. Maybe just convert PascalCase -> WikiCase in the docs?

I'd be curious if the common-case for 64 bit integers is that they're used as identifiers or actually used for math. If for math, it seems important we resolve the...