> I appreciate this approach is additional work and pulls in another > "technology" that isn't a natural fit in the ACME ecosystem.
I think using CDDL is a bad idea. As you point out, it's not a natural fit. I looked at Appendix B of RFC 8610, and while I *think* it would work, I'm not positive. None of the other ACME documents have used a schema and seem to be acceptable. If the WG authors really think a schema language is needed, I betcha they could craft ABNF or even ASN.1 (ISO X.697 if you need to go that far). Make Appendix B informative and change the second bullet in 5.6 to be "A description of the extension syntax." Beware of over-specifying. When JSON Schema finally becomes published, re-open ACME (heh:) and put out an "updates" document that makes everything like you want it to be. >Also, there are discussions > between the leaders of the JSON Schema effort and people on the HTTP-API > working group, with the goal of standardizing it there. As a co-chair of that group I'll say that the HTTP-API group does not feel json schema belongs there as we have too much work already and JSON isn't just about API's. My guess is it will end up in another group. Which will of course mean things take even longer. _______________________________________________ Acme mailing list [email protected] https://www.ietf.org/mailman/listinfo/acme
