Rico Suter

/cc @lahma @0xced We need to fix this ASAP... major breaking change for non-latest .NET users. I think we need to introduce CSharpVersion configuration property (same as TypeScriptVersion property, set...

> Or emit a `#if NET9_0_OR_GREATER` in the template yeah good idea, simple and does not need yet another config :-), any way this could introduce another breaking change? can...

... ill have a look

> Also, yes, the new JsonStringEnumMemberName attribute is needed to properly serialize an enum value that does not match the enum name exactly. The System.Runtime.Serialization.EnumMember has never been supported by...

Will roll this back and introduce a JsonLibraryVersion instead... This worked because NSwag generated a custom JSON converter which actually considered this EnumValue attribute and all worked... Now we have...

Fix here: https://github.com/RicoSuter/NJsonSchema/pull/1865

Yup, I think it was introduced for a simpler template - but I agree if it's always false (probably when InjectHttpClient = true) then we should not generate that at...

This is essentially a wrong spec, ie the type must be "integer" https://github.com/RicoSuter/NJsonSchema/blob/master/src/NJsonSchema/JsonObjectType.cs#L32 Not sure whether we should handle this "wrong" spec in your expected way: https://github.com/RicoSuter/NJsonSchema/blob/master/src/NJsonSchema.CodeGeneration.CSharp/CSharpTypeResolver.cs#L97 and here https://github.com/RicoSuter/NJsonSchema/blob/master/src/NJsonSchema.CodeGeneration.TypeScript/TypeScriptTypeResolver.cs#L106

> If they don’t want to change the spec, is there any recommended workarounds? Preprocess the spec so that it fits the generator expectations...

I have the feeling that this is just a "wrong configuration" and this PR breaks some config permutations.