RE: Thinking differently about the site local problem (was: RE: site local addresses (was Re: Fw: Welcome to the InterNAT...))

["Jeroen Massar" <[e-mail]>]

Margaret Wasserman [mailto:[e-mail]] wrote:

> Hi Jeroen,
> 
> > > >Most OS's require a (unique) hostname to be entered/automatically
> > > >generated on install
> > > >
> > > False.
> >
> >And is there any reasoned argument instead of the simple 'false'?
> 
> Some examples certainly would have been helpful...
> 
> I can give one common example.  Special-purpose client systems
> (consumer devices, car infotainment systems, etc.) seldom require
> or configure a DNS name.
> 
> In most IPv4 situations, though, these types of systems would
> receive their addresses through DHCP, usually through an ISPs
> DHCP server.  In many cases, ISPs provide generic (e.g.
> hostNNNNNNN.isp.com) host names for all of the addresses in their
> DHCP pool, so that reverse lookups will work, etc.
> 
> But the operating system and TCP/IP stack on the device do
> not actually require that a hostname be configured.

My thought went out from both the manual page of a typical Unix box:
8<-------
 Hostname is used to either set or display the current
 host or domain name of the system. This name is used
 by many of the networking programs to identify the
 machine. The domain name is also used by NIS/YP.
------->8

And the fact that windows boxes nowadays also
autogenerate a hostname based on some of the
properties of the machine to overcome clashes
in the SMB/CIFS namespace. Note that this still
has nothing to with IP.

Phones and other devices that could be used
for transfering files usually autogenerate a
hostname based on their cellphonenumber.

But indeed there are bound to be apparatus which 
don't have such a feature, thus one will have to
type the full IP then if you want to do something.
Fortunatly finding out all the hosts on a subnet
isn't that trivial anymore ff02::1 ;)
But I wonder how consumer friendly that is...

Greets,
 Jeroen