On Wed, Jan 3, 2018 at 7:18 AM, Stephen Farrell <stephen.farr...@cs.tcd.ie> wrote: > [...] the capport wg - I'd guess folks > there are more aware of the full range of cases that may need > handling and of how to try interpose the portal web page stuff > before other applications see the n/w as active (or whatever it > is they're doing with HTTP:-).
As chair of capport, this is definitely something for that group. The current approach we are taking there avoids having user equipment attempt to connect to anything at all, which avoids this class of problem. If you look at modern devices, they all probe a network before making the interface available to applications, and most of what we would be doing exists at that probing phase. We are additionally considering a network-based signal for those cases where the attempt is made anyway, but I don't think that is firm yet. Captive portal cases are separate to those related to selective blocking of names or destinations, which is essentially a censorship mechanism. We've been careful to avoid creating mechanisms to support that sort of discrimination. _______________________________________________ TLS mailing list TLS@ietf.org https://www.ietf.org/mailman/listinfo/tls
