Re: [SLUG] ipv6

[Holger Zuleger <Holger.Zuleger@xxxxxxxx>]

Anand Kumria wrote:
> On Mon, Dec 19, 2005 at 10:30:43PM +1100, Visser, Martin wrote:
>
> > Anand Kumria wrote :-
> >
> >
> > > The site-local prefix (fe80) has been deprecated (rfc3879), instead you
> >
> > want IPv6 local addresses (rfc4193) which you
> >
> > > can self-generate with tools such as:
> >
> > http://www.hznet.de/tools/generate-uniq-local-ipv6-unicast-addr.sh
> >  
> >
> > Hmm, I dropped off the IETF announce lists a few years ago so I have
> > missed this fairly significant change. One thing I noticed though was
> > that this script doesn't comply with the mentioned RFC. I am guessing it
> > may have been written against an earlier draft (yep, the script says
> > Sept. 2004). 
>
>
> Hmm, I thought I had updated the URL before I sent out my email.
>
> Anyway, http://www.hznet.de/tools/generate-rfc4193-addr.sh is one
> which seems to comply with the RFC.  It might be worthwhile checking
> this as throughly as you have the last one though.
>
>
> > While it creates a pseudo-random address, a few problems I see are that
> > it uses FD00::/8 as the prefix (instead of FC00::/7 which means it only
> > tries to use half of the available space) and MD5 instead of SHA1 as the
No! This is what RFC 4193 call a "locally" assigned uniq ipv6 unicast 
address. In one of the former drafts the addresses with an L bit set to 
0 are called "globally" assigned local ipv6 unicast addresses.
Currently these are not defined (See Chapter 3.2).

> > digest/randomizer. I know I am pedantic but one of the assumptions in
> > this RFC (section 3.2.1) is that all generators of locally assigned
> > global IDs use the same algorithm.
>
>
> Holger, it is probably worthwhile if you generate a 303 - See
> other  response if someone tries to access the earlier shell script.
Yes, done and thank you for the hint.

> Regards,
> Anand

Greets
Holger