[Asrg] Re: Another draft of the DNSxL document

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

Bill Cole wrote:

>> The example still uses 127.0.0.1.  It shouldn't, that IP is
>> never listed

> False to fact.

Okay, add "normally", that's what I meant.  A DNSBL of private
and reserved IPs is somewhat "special".  Testing one I know:

2.0.0.127.combined-hib.dnsiplists.completewhois.com = 127.0.0.2

host: unknown host 1.0.0.127.combined-hib.dnsiplists.completewhois.com

For a list with 127.0.0.1 as potential entry 127.0.0.2 is most
probably no test entry, so that's IMHO really a "special" case,
where no "normal" conventions work as expected.

Was it Osirusoft, that ended with "all IPs listed" ?  A sanity
test "127.0.0.1 not listed" could catch this condition.

>> and never returned as result.
> I'm not sure 'never' is really true.

I'm sure, it's also only "normally" :-)  A counter-example is
2.0.0.127.wadb.isipp.com = 127.0.0.1

> It is never returned wisely, certainly.

Maybe the 127.0.0.1/31 issue should be added to the "security
considerations" (?)

> It has been done intentionally by list maintainers as a test
> record, without adequate appreciation for the depth of
> cluelessness among their users.

Yes, that's what I wanted to express by "never" :-)  The "fix"
s/1+2=3/2+4=6/ would replace the 127.0.0.1 example by 127.0.0.4
avoiding the critical bit 0.
                             Bye, Frank



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