It is not hard to see that e.g., NULL encryption violates the properties. Sure. And for years we thought MD5 met the properties, until it didn’t. And now, RSA meets the properties, until it doesn’t.
The alternative is that someone afterwards need to write a standards track draft and progress that through IETF. As an author of such a draft I would rather not have do that work. I would much rather help evaluating if an item violates the properties before registration. That’s better than trusting security to a handful of people. I mean, if you’re making a judgement that global security needs to move away from an algorithm, having to get a document through standards track seems a very small price to pay. I don’t want that job, and I’d quit if the TLS registries were changed that way. I don’t think it’s likely.
_______________________________________________ TLS mailing list [email protected] https://www.ietf.org/mailman/listinfo/tls
