> On Apr 12, 2018, at 9:07 PM, Martin Thomson <martin.thom...@gmail.com> wrote: > > On Fri, Apr 13, 2018 at 1:55 PM, Christopher Wood > <christopherwoo...@gmail.com> wrote: >> Yes — we’re currently working on an I-D that would use the context for >> “special” tickets. Depending on where that goes, if anywhere, we may or may >> not need to keep the context. As you suggest, distinguishing between >> responses and spurious NSTs doesn’t *seem* like a problem. > > Maybe the right way to deal with this is to put an extensions block in > the request. Then you only have to resolve the question of whether > NST answers the ClientHello or this new message...’
Indeed. That might work just as well, if not better. In any case, as written right now, we prohibit servers from sending spurious NSTs if both parties negotiate request support. If implementations honor that requirement, we won’t have a distinguishing problem. We then need only decide what is the best encoding strategy for request identity/context, if desired. _______________________________________________ TLS mailing list TLS@ietf.org https://www.ietf.org/mailman/listinfo/tls
