Re: [DNA] DNA and TSLLAO

[James Kempf <kempf@docomolabs-usa.com>]

I think TSLLAO needs to be folded into the DNA document. I think we should
maintain the tradition of defining enhancements to ND based on a link
techology with a host MAC address that was started by 2461, and therefore in
the general case the router will need the TSLLAO in order to unicast the RA
back to host. Other link layer types or corner-case uses of 802 link
technologies that don't support MAC addresses on the wireless link (like
802.16) need to be seen as exceptions, to be handled by extension RFCs (like
RFC 2491 does for ATM).

            jak

----- Original Message ----- 
From: "Greg Daley" <greg.daley@eng.monash.edu.au>
To: <dna@eng.monash.edu.au>
Cc: "Pekka Nikander" <pekka.nikander@nomadiclab.com>
Sent: Sunday, June 26, 2005 10:53 PM
Subject: [DNA] DNA and TSLLAO


> Dear WG,
>
> Please be aware that there's some conflict of interest here
> (being one of the authors of the document), so I'd like to
> be guided by the WG's opinion.
>
> I've been discussing the use of
> Tentative Source Link-Layer Address Options (TSLLAO)s
> with some of the people in the DT, and it seems that
> there's some belief that TSLLAOs are useful for DNA.
>
> I'd like to confirm if this is the case on the mailing list,
> and if so, determine what to do with the existing draft.
>
> The draft is: draft-daley-ipv6-tsllao-01.txt
>
> It's referenced from draft-pentland-dna-protocol-00.txt,
> and is relied upon for delivery of unicast fast RAs to
> Optimistic addresses.
>
> There's an experimental implementation of the technology
> available using RADVD, as well as a partial implementation
> (doesn't use optimistic DAD) for an RS sender.
>
> Do people think it would be useful for this technology to
> be standardized?
>
> If so, where should it go:
>
> a) within the DNA solution
>
> b) as a separate document within DNA
>
> c) ask IPv6 (or another suggested WG) to work on it?
>
> At this stage, discussion is encouraged, so that people
> can devlop opinions over time.
>
> If there's significant interest, though, I can tentatively
> approach the IPv6 chairs for their (early) opinion too.
>
> Please tell me what you think, and alternative ideas and
> opinions are welcome.
>
> Greg.