Re: [mif] declaring interface 'up', with WiFi DNS/HTTP interception (login) proxies [was RE: DNS selection with HE-MIF]

Le 2013-02-08 à 08:48, Michael Richardson a écrit :

> 
>>>>>> "Keith" == Keith Moore <moore@network-heretics.com> writes:
>>> For MIF -- not just HE-MIF, but all of MIF -- we should not
>>> declare an interface "up" until such a validation succeeds.  It
>>> is unfortunate this is not solved at layer 2, where it arguably
>>> belongs.
> 
>    Keith> Would it be worthwhile for MIF to start making a list of
>    Keith> things that really need solutions elsewhere?  Even if there
>    Keith> are hacks or heuristics that are used in the absence of such
>    Keith> solutions?
> 
> Yes.
> 
> In the portal case, we need a DHCP "login required" message.

specially in more open portals, IPv6 deployments may use RA, so it has to also involve RA.

Marc.

> It would be nice if we also had a BCP on how to signal and upgrade
> From HTTP login to some DHCP EAP, perhaps using a EAP-TLS resume 
> From the HTTP session state.  This would permit captive portals to 
> recognize re-logins.
> 
> -- 
> Michael Richardson <mcr+IETF@sandelman.ca>, Sandelman Software Works 
> 
> 
> _______________________________________________
> mif mailing list
> mif@ietf.org
> https://www.ietf.org/mailman/listinfo/mif