Matthew Pope

This was fixed serendipitously as part of #238. The only action required is to add the repro test case as a regression test.

Using the [`plusType()`](https://github.com/amazon-ion/ion-schema-kotlin/blob/bbab678c12fdd07bcb9b5d632fb7364a04541a73/ion-schema/src/main/kotlin/com/amazon/ionschema/Schema.kt#L100-L106) method, you can create a new copy of the schema that has the new type added to it.

I don't think it's a good idea to change the behavior of the existing `newType()` API, but I definitely think it could be useful to have a `plusTypes()` API that...

Thanks for this example. You've helped identify an edge case here where a schema with only one top-level Ion value is not being handled correctly. We'll have a fix for...

I'd also like to understand—do you need to be adding the types one by one or could you do something like this? ``` String islString = "type::{ name: foo, type:...

Thanks for reaching out! > Not sure why when this could have been easy to be allowed. Ion Schema Kotlin does not allow adding arbitrary annotations because the Ion Schema...

> I have a group of optional fields that either must all be present and valid, or none of them can be present. Here is a cookbook describing how to...

Both of these are a little verbose. We do have an open issue to improve this https://github.com/amazon-ion/ion-schema/issues/113. If the proposed solution here is adopted, then it would become much simpler....

Also created https://github.com/amzn/ion-schema/issues/38 to address the ambiguity in the spec around `nullable::`.

Do you think this is an issue in the implementation, or a mistake in the spec? With the existing `AuthorityFilesystem`, you can't actually load any schema with an id of...