Re: [v6ops] Please review the No IPv4 draft

[Nick Hilliard <[e-mail]>]

On 15/04/2014 04:12, Ted Lemon wrote:
> Because you are changing the protocol state machine.

it's a single new state with a single transition point, and it would need
to added to the state machine regardless of whether the signal comes
directly through dhcpv4 or indirectly through dhcpv6 or ra.  The difference
between the two is that dhcpv6/ra would need an asynchronous transition
which significantly complicates the dhcpv4 FSM because you would now have
the possibility of a transition from any dhcpv4 state to the No IPv4 state.
 IOW, you're creating a longjmp style exception mechanism in the FSM.  This
is not a sensible thing to aim for.

Nick