Date: 03 Feb 1999 08:42:50 -0500 From: Lowell Gilbert <lowell@world.std.com> To: freebsd-chat@FreeBSD.ORG, tlambert@primenet.com Subject: Re: DHCP Autoconfig (Was: Lets Endorse KDE) Message-ID: <rd67ltzwqlh.fsf@world.std.com> In-Reply-To: Terry Lambert's message of Wed, 3 Feb 1999 03:24:46 %2B0000 (GMT) References: <199902030324.UAA04247@usr02.primenet.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Terry Lambert <tlambert@primenet.com> writes:
> > If I'm not mistaken, the autoconfiguration process checks the address it
> > picks before it sticks with it. From the IETF Internet Draft on the
> > subject:
> >
> > Once a DHCP Client has determined it must auto-configure an IP
> > address, it chooses an address. The algorithm for choosing an
> > address is implementation dependant. The address range to use MUST
> > be "169.254/16", which is registered with the IANA as the LINKLOCAL
> > net.
>
> Is this a draft from Microsoft, or is it from people who don't
> violate protocol layering with things like PAP?
>
> Do you have a reference to the draft?
draft-ietf-dhc-autoconfig-03.txt appears to be the current version.
> Also, if it's not MS, it would be just like them to declare that if
> there isn't a DHCP server, then they aren't a DHCP client, and they
> will therefore do what they damn well please. 8-(. Makes you hope
> it's their draft...
I'm not sure what Terry's objection is here. DHCP *already* expects
hosts to check the address they get from the server before using it,
through exactly the same mechanism (ARP where appropriate). It's
tempting to break layering to implement that address check, but it
isn't necessary.
The draft was written by a networking staffer at Carnegie Mellon
University, and he credits both Microsoft and Apple. I really don't
see the relevance of the corporate connection, unless we're going to
develop a Microsoft-specific version of NIH -- an affliction I haven't
ascribed to Terry Lambert.
The idea behind this mechanism is pretty much the same as stateless
autoconfiguration in IPv6, although it's a little less useful because
you can't assume everyone on your own link uses it. The main reason I
can see for worrying about it is that (unlike RFC1918) there are a
*lot* of routers that don't know those addresses are supposed to be
unroutable.
Be well.
Lowell
Unfortunately, I've fallen completely out of touch with the DHC
working group. I had long been advocating that they declare victory
and go home (at least with regards to IPv4), and occasionally came
back from IETF meeting feeling like "Animal," after the credits of
"The Muppet Movie," jumping up and down yelling "Go home! Go home!"
When some people started insisting on adding security to DHCP, I
decided that adding a jet engine to that particular horse cart was too
silly to even pay attention to.
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-chat" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?rd67ltzwqlh.fsf>