Re: [BEHAVE] [Technical Errata Reported] RFC6052 (5547)

[Christian Huitema <[e-mail]>]

To be clear, I am not trying to assert that RFC 6052 is technically
perfect. There may very well be good reasons to write an RFC 6052bis. We
have now several years of deployment experience, and we may possibly
have learned a thing or two. I am just saying that the errata process is
the wrong way to do that.

-- Christian Huitema


On 11/8/2018 6:00 AM, [e-mail] wrote:
> Hi all, 
>
> I echo what Christian said.
>
> Please note that if one wants to relax the constraint about private IPv4 address in 6052, the WKP prefix defined in https://tools.ietf.org/html/rfc8215 (which Fred is surely familiar with) can be used. 
>
> Cheers,
> Med
>
>> -----Message d'origine-----
>> De : Behave [mailto:[e-mail]] De la part de Christian Huitema
>> Envoyé : mardi 6 novembre 2018 11:18
>> À : Fred Baker; RFC Errata System
>> Cc : Congxiao Bao; [e-mail]; Dave Thaler; Xing
>> Li; [e-mail]; [e-mail]; [e-mail];
>> [e-mail]; [e-mail]
>> Objet : Re: [BEHAVE] [Technical Errata Reported] RFC6052 (5547)
>>
>>
>>
>> On 11/6/2018 4:23 PM, Fred Baker wrote:
>>> On second thought, the exact case is incorrect. I still think it's a silly
>> restriction, but the private address in the case would not be placed into the
>> Well-Known Prefix (and used as a destination address); it would be put into a
>> /64 out of the subscriber's prefix (and used as a source address). So I can
>> live without this change.
>>
>> Good.
>>
>> I think that the text should not be changed by an errata. The text is a
>> fair rendering of our technical assessment at the time. We did not want
>> to use the WKP to build ambiguous IPv6 addresses, and wanted to make
>> sure that private IPv4 addresses will be mapped using private prefixes.
>> It may or may not have been the right decision, but this is definitely
>> what we intended.
>>
>> -- Christian Huitema