Milter-greylist

Hi,

>>>>> On Fri, 28 Jul 2006 22:54:51 +0200
>>>>> manu@... said:

manu> Hajimu UMEMOTO <ume@...> wrote:

> Okay.  Then, I think we have two choices:
> 
>   1) disable DNSRBL query against an IPv6 address
>   2) make our format for an IPv6 address
> 
> If we choose 2), I offer to use similar format with IPv6 PTR RR like:
> 
>   0.1.2.3.4.5.6.7.8.9.a.b.c.d.e.f.0.1.2.3.4.5.6.7.8.9.a.b.c.d.e.f.example.com.
> 
> If you like it, I'll implement it.

manu> Well, what's the point of making a format, if we have no garantee that
manu> people will follow the same path? I'd be in favor of handling IPv6 the
manu> day it becomes useful and specified.

manu> What do you think?

AFAIK, there is no standard nor de facto to handle an IPv6 address by
DNSRBL.  Further, there are no available service which supports an
IPv6 address.  However, if someone want to serve the server, it will
be possible, as far as there is certain format for an IPv6 address.
In current implementation, the form is

	inet_ntop(reverse(address)) + domain

even for an IPv6 address.  Since inet_ntop produces a canonical form
of IPv6 address representation, it is hard to specify A RR in their
zone file.  Further, it contains `:' rather than `.'.  This is why I
offer an IPv6 PTR RR like format.
Even if it will not be a de facto, still we will be able to use this
format in anyway.  However, if DNSRBL will incorporate another format
in their service, we will need to chase the format.  Perhaps, we can
support both format with a configure option.
If we don't want to have a compatibility issue in the future, we
should disable a DNSRBL query against an IPv6 address for now.

Sincerely,

--
Hajimu UMEMOTO @ Internet Mutual Aid Society Yokohama, Japan
ume@...  ume@{,jp.}FreeBSD.org
http://www.imasy.org/~ume/