Bart Koelman

Hi @baywet, thanks for your response. > I'm not super familiar with JSON API, but from what I understand query parameters can be separated in two main categories Unfortunately, it's...

> Probably not, here is the rationale for it... That makes sense. You've convinced me that's the better decision for Kiota. Is there a way to plug this into Kiota...

What I forgot to mention is that JsonApiDotNetCore has a setting to allow unknown query string parameters to be passed through. When enabled (false by default), API developers can implement...

> Question about the hooks: are they registered somewhere? meaning do you have a way to enumerate them so they could also be projected to query parameters? Unfortunately, no. It...

Created https://github.com/microsoft/OpenAPI.NET/issues/1489.

Not anytime soon, I'm pretty occupied with other work at the moment.

> I like the solution proposed by @markm77 where, if I understood well, and assuming we don't have any switch, we'd only make as non-nullable properties/return types that are required...

Related to this, NSwag goes beyond just scalar properties. I've seen cases where it generates: ```c# public class Customer { [Newtonsoft.Json.JsonProperty( ... )] public CustomerAddress Address { get; set; }...

> I don't think you're suggesting taking a dependency on newtonsoft annotations That's true. It's just an implementation detail in NSwag to make Newtonsoft do the right thing. It would...

> The primary problem is that the use of required/nullable in OpenAPI is broken. I disagree. It is well-defined, just not so convenient for simplistic unstructured APIs, like the default...