On Thu, Sep 21, 2023 at 6:55 AM Pawel Kowalik <kowa...@denic.de> wrote: > > 4. Schema languages give a false sense of conformance. Conformance > tools are far more important. > > Schema won't replace the conformance tools for sure, as not every constraint > or relation can be covered with schema. But it makes it a lot of easier for > the implementer to have a solid base for a conforming implementation.
It is true that DDL can make things easier, especially for code generation to start creating a codebase. But it is certainly not enough as the industry found out with SOAP and UDDI. A number of years ago, Pete Cordell and I put a good bit of thought into this: https://datatracker.ietf.org/doc/html/draft-newton-json-content-rules-09 https://datatracker.ietf.org/doc/draft-cordell-jcr-co-constraints > Therefore the default position became "use CDDL". > > Tooling for CDDL is like almost non existent and JSON seems to be 2nd class > citizen in this specification. Do you really believe it would have a take off > outside of CBOR? That was never my position, but it does seem to be the prevailing attitude of the IETF. Maybe with time we can now show CDDL was not the answer for JSON. -andy _______________________________________________ regext mailing list regext@ietf.org https://www.ietf.org/mailman/listinfo/regext
