Re: [v6ops] New Version Notification for draft-ipversion6-loopback-prefix-00.txt

[joel jaeggli <[e-mail]>]

I'm not going to comment on the merits of this spefic proposal at this
time. one thing catches my however which is that mappings between v4 and
v6 addresses used for the same  purpose make very a little sense 7f::
doesn't have a meaning becasue of some magical congruence with 127
decimal and makes no particular sense in the absence of v4.

although this draft contends with a particular aspect of mapping between
v4 and v6 and is sadly expired I think it's instructive as to intuiting
meaning from equivalence:

https://tools.ietf.org/html/draft-itojun-v6ops-v4mapped-harmful-02

Beyond that  in general we should leave the actual assignment of special
use-prefixes to iana as stewards, we request that they be allocated.
they assign them.


On 2/17/15 6:55 AM, Edward Lewis wrote:
> On 2/17/15, 7:43, "Eric Vyncke (evyncke)" <[e-mail]> wrote:
> 
>> Indeed. Clarity always wins, this is why I tend to like this I-D (albeit
>> it should be rewritten).
>>
>> BTW, the choice of ::/64 is probably not the best choice as it includes
>> the deprecated ::/96 for IPv4-compatible IPv6 address.
> 
> I agree on both counts (needs rewriting and a different block is needed).
> 
> What I’m thinking of writing is a draft to set aside 7F::/64 (because 0x7F
> is 127 in decimal) and set:
> 
> Source - - - - - - - - False
> Destination  - - - - - False
> Forwardable  - - - - - False
> Global - - - - - - - - False
> Reserved-by-Protocol - True
> 
> 
> These are the same values as for 127.0.0.0/8 in RFC 6890.
> 
> The question remains for me, whether the Protocol should make this a
> loopback or refuse to create the datagram.  By “make this a loopback” I
> mean specify that one of the addresses should by default be ‘ifconfiged’
> to the loopback interface and the -prefixlen 64 route be added too.
> 
> 
> 
> _______________________________________________
> v6ops mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/v6ops
>