paul@whooppee.com (Paul Goyette) writes:
> Now, I _know_ that the zeroth host is kinda special, but my new ISP has
> me set up this way:

You should be able to make this work.  This is also a good test to see
if there is any "dirty" code that makes silly assumptions about
host-0.  That should be limited to RIP-v1 (no longer a valid RFC
standard) and paleolithic versions of OSPF.  Everything else
*hopefully* accepts a netmask so that it can differentiate between
10.1.2.0/24 and 10.1.2.0/32.

As a lark I went through some dialup router code a while back and
fixed it up so that it would run with host-0's.  It worked fine once I
made sure all the routines passed along address *and* netmask.

> Is there any hope for me?  Is there anything I can do to make ARP handle
> the zeroth host?

Find that spot in the arp code and beat it bloody?

I assume someone put the host-0 hack in there in deference to the
ancient 4.1BSD (???) practice of having the 0th-address be the
broadcast.  Perhaps its time to think about having a switch to turn
this nonsense off?  Lots of ISP's are handing out 4-host and 8-host
subnets and blowing away two addresses (0, -1) gets to be kinda
expensive.

-wolfgang
-- 
       Wolfgang Rupprecht <wolfgang+gnus@dailyplanet.wsrcc.com>
		    http://www.wsrcc.com/wolfgang/
DGPS signals via the Internet  http://www.wsrcc.com/wolfgang/gps/dgps-ip.html