|
> I'll rev the DTLS document soon. We can then see how much of it > simply references radsec. If the DTLS-specific portions are small, it > may be good to add them to the radsec document. Even if the DTLS-specific portions are small, there is still a "compliance" issue. That is, if two vendors say "I implement RADSEC" -- what would it mean, if one could be doing RADIUS over TLS/TCP and another, RADIUS over DTLS/UDP? Whereas if there are two separate documents, and someone says "I implement RADIUS over DTLS", now we know what transport they are using. There are similar concerns about RADIUS over TCP, but we discussed indicating in the document that this is only intended for use with RADSEC. |