Version 1's way of handling validation is efficient. Leave it up to the other methods to handle anything that is not validation.

I had forgotten that how the values are decoded will affect the schema validation process. As with the previous version of the library, the developer will be required to provide the decoded value as they wanted it.