[Asrg] Proposed fix for 127.0.0.1 (was: New draft of the DNSBL definition document)

[Frank Ellermann <nobody at xyzzy.claranet.de>]

Todd Vierling wrote:

> My gut-feeling guess would be that it was ideal due to the
> guarantee that it never really exists as a usable domain.

Yes, but the RfC reserving example.com also reserves test, so
there would be a more obvious choice.  For some time IANA had
an "example Web site" at example.cno, better stay away from it.

>From my POV three hould be three rules for all BLs:

- 127.0.0.2 is always listed (for a RHSBL read 2.0.0.127)
- 127.0.0.1 is never listed (for a RHSBL read 1.0.0.127)
- a "listed" result is 127.0.0.0/8 excluding 127.0.0.0/31
- for bitmapped results that excludes "odd" IPs 127.0.0.1,
  127.0.0.3 etc. up to 127.0.0.255, but OPM style is okay:
  127.1.0.1 etc. won't be confused with 127.0.0.1

For bitmapped results it could be an idea to include the
version number of the bit definitions in the result:
127,nn.hi.lo, nn = version 1..255, and hi.lo for 16 sets.

But that's only an idea, I don't know any RHSBL with more
than 7 sets, SURBL has 6 sets at the moment (but it won't
touch bit 0 when it gets more sources, 127.0.0.1 is taboo).

Back to the draft, here's a fix for the 127.0.0.1 issue:

=== old ===
example, the bit masks for the two sublists might be 127.0.0.1
and 127.0.0.2, in which case an entry for an IP on both lists
would be 127.0.0.3.  With multiple A records, each sublist has
=== new ===
example, the bit masks for the two sublists might be 127.0.0.2
and 127.0.0.4, in which case an entry for an IP on both lists
would be 127.0.0.6.  With multiple A records, each sublist has
=== end ===
                          Bye, Frank



_______________________________________________
Asrg mailing list
Asrg at ietf.org
https://www1.ietf.org/mailman/listinfo/asrg