Re: [dhcwg] draft-ietf-dhc-dhcpv6-opt-nisconfig-05.txt

[Stig Venaas <[e-mail]>]

On Tue, Feb 10, 2004 at 02:43:27PM +0900, JINMEI Tatuya / ?$B?@L@C#:H wrote:
[...]
> - if we really go with this approach, we should also consider
>   draft-itojun-v6ops-v4mapped-harmful-02.txt
>   which proposes "nodes SHOULD NOT generate packets that contain
>   IPv4-mapped addresses in any field" in a packet sent on the wire in
>   order to avoid confusion with API usages of IPv4-mapped addresses.
>   Though this is still an individual draft, there seems to be a
>   consensus on "no IPv4-mapped IPv6 addresses in the wire", at least
>   to some extent.

I agree with the concerns for IP header fields which I believe the
draft discusses. I don't see immediate problems with this usage in
the IP payload. I don't think the draft says it's wrong to use them
in payload either. It says something about no IPv4-mapped addresses
on the wire, but when discussing issues and other places, it only
talks about IP header fields, which I think is the problem it
discusses, and rightly so.

So I think the concensus is "no IPv4-mapped IPv6 addresses" in IPv6
header fields. Perhaps my understanding is different from the rest.
But I suggest people read the draft before refusing this approach.

Stig

_______________________________________________
dhcwg mailing list
[e-mail]
https://www1.ietf.org/mailman/listinfo/dhcwg