[2005] in linux-net channel archive
Re: Dynamic IPs
daemon@ATHENA.MIT.EDU (Dan Merillat)
Mon Mar 4 23:20:22 1996
Date: Mon, 4 Mar 1996 19:56:07 -0500 (EST)
From: Dan Merillat <Dan.Merillat@ao.net>
To: Ken Yap <ken@syd.dit.csiro.au>
cc: Linux-Net List <linux-net@vger.rutgers.edu>
In-Reply-To: <199603042240.IAA14459@alba.syd.dit.CSIRO.AU>
On Tue, 5 Mar 1996, Ken Yap wrote:
> Date: Tue, 05 Mar 1996 09:39:46 +1100
> From: Ken Yap <ken@syd.dit.csiro.au>
> To: Dan Merillat <Dan.Merillat@ao.net>
> Subject: Re: Dynamic IPs
>
> > While on the subject, how hard would it be to integrate the new standard
> > (when it becomes standard) into the kernel/existing apps? Obviously,
> > stuff that deals with the net would have to be recompiled, but MUA's etc
> > that just deal with hostnames, would not.
>
> I'm not sure of the status but I believe it's approved draft and
> several implementations are under way.
>
> There will be some new APIs to handle the new addresses and some new
> notation. Instead of dotted decimal, it will be coloned hex :-) with
> some backward compatibility for v4 addresses, i.e. ::192.168.1.1.
Well, ::ffff:x.x.x.x
Aiee! Ugly!
And I CAN memorise dotted quad, but this ipv6 is EVIL to try to memorize.
Of course, since it's hex, we can pick cool ip addresses:
ACAD:ACE:......
^^^^ 0 padded....
or DAD:ADD:AFEC::BAD:DAD:CAD:DEAD
Looks valid here. :-)
A few companys luck out too...
DEC, ABC, AAA, ACAD....
Any reason why those wouldn't be valid? And are machines _REALLY_ getting
the full 128 if you specify their ipv6 ip address? I.E. testing DNS is
going to SUCK!
--Dan
