Working on howl port

[Andrea Campi]

On Sat, Dec 11, 2004 at 03:56:33PM -0500, Chuck Swiger wrote:
> >Just to check my assumptions: is it reasonable to assume autoipd
> >has total control over the 169.254 block? I don't want to have to
> >bother about preserving any existing address in that range etc.
> 
> No, it is not reasonable.  Although if you make that assumption for the 
> first implementation, it's probably mostly harmless again.  However, 

Ah no, I meant on the local machine. What I mean is that by running
zeroconf the user is giving total control of that block, and in
particular is going to avoid explicitely configuring an interface
with such an address. I figure that's quite reasonable.

> Anyway, perhaps my responses would be more helpful if you explained the use 
> case you are trying to get going, first?

Unless I'm mistaken, using SIOCAIFADDR instead of SIOCSIFADDR means
that I do need to bother about other addresses in the same network
on the same or on another interface. My plan is to replace any
such address as needed, as long as I can trust they are not of any
special meaning to the user.

Additionally, if I can expect the user not to set a 169.254 address,
I can assume RTM_ADD/RTM_NEWADDR messages for such an address come
from dhclient failing to get a lease. Think of this scenario:

 - unconnected laptop on a bus, I get an addess by zeroconf;
 - I arrive at the office and the DHCP server gives out an address,
zeroconf goes to sleep;
 - the DHCP server dies and come renew time dhclient notices and
sets a 169.254 address.

I'd like to be able to notice this case and decide it's really
dhclient trying to tell me something, and not have to worry about
a user trying to set a random IP address.


Anyway, thinking again, this is probably less important than I
first thought. I'm mostly thinking out aloud here. It just
occurred to me that the second part can probably be better
dealt with by having an useful set of scripts for dhclient.


I think I should now get some more steps and expecially tests
done, and come back when I have a clearer idea of how they
interact. I'll still post some interim steps for feedback on
the implementation.


Bye,
	Andrea


-- 
            Give a man a fish and you feed him for a day;
     teach him to use the Net and he won't bother you for weeks.