RFC 4291

[Frank Ellermann]

Hi, somehow RFC 4408 managed to have a normative reference
to RFC 3513 when that was already obsoleted by RFC 4291.

The latter was updated by RFCs 5952 and 6052. RFC 5952
is relevant for SPF, because it fixes some details for the
textual representation of IPv6 addresses. Section 3.3.3
is also interesting for ARF.

For RFC 6052 I'm not sure if it is relevant for SPF, what
do the IPv6 experts here think? We knew ::FFFF:0:0/96 for
IPv4-mapped IPv6 addresses. IIRC it is used in the test
suite and/or in the reference implementation.

Unsurprisingly we did not know RFC 6052 64:ff9b::/96 for
"IPv4-embedded IPv6 addresses with the well-known prefix"
(that long name is used in RFC 6052). I think that this
is irrelevant for SPF implementations, but I'm very far
from sure.

-Frank

[Scott Kitterman]

Post by Frank Ellermann
Hi, somehow RFC 4408 managed to have a normative reference
to RFC 3513 when that was already obsoleted by RFC 4291.
The latter was updated by RFCs 5952 and 6052. RFC 5952
is relevant for SPF, because it fixes some details for the
textual representation of IPv6 addresses. Section 3.3.3
is also interesting for ARF.
For RFC 6052 I'm not sure if it is relevant for SPF, what
do the IPv6 experts here think? We knew ::FFFF:0:0/96 for
IPv4-mapped IPv6 addresses. IIRC it is used in the test
suite and/or in the reference implementation.
Unsurprisingly we did not know RFC 6052 64:ff9b::/96 for
"IPv4-embedded IPv6 addresses with the well-known prefix"
(that long name is used in RFC 6052). I think that this
is irrelevant for SPF implementations, but I'm very far
from sure.

My initial reaction is to update the reference and not touch the text from
4408. You can't implement IPv6 address parsing without understanding how it
works and 4408bis is not the place to explain it. We don't treat 169.254.x.x
specially for IPv4, so I don't think we need to deal with "well-known prefix"
specially either.

Scott K

[Murray S. Kucherawy]

-----Original Message-----
Sent: Friday, September 09, 2011 6:28 AM
Subject: Re: [spf-discuss] RFC 4291
My initial reaction is to update the reference and not touch the text from
4408. You can't implement IPv6 address parsing without understanding how it
works and 4408bis is not the place to explain it. We don't treat 169.254.x.x
specially for IPv4, so I don't think we need to deal with "well-known prefix"
specially either.

Much as I hate "me too" emails:

Me too.

-MSK