On 29 Jan 2004, at 14:34, Ulrich Kiermayr wrote:
1. Role lacks of the two pgp-attributes. Might it be worth to put them as optional into that templates? [I'd say probably yes, because it might be useful in other situations as well]
I'ld agree. Safety options should always be available.
2. Protection against 'illegal' referecing. (Meaning "Hey UCD has a nice role for abuse, i reference this role in my inetnums as well") There we had a discussion in the context of the org object, putting a generalized machinery (mnt-ref, ref-nfy) in place. IMHO this _should_ be added (optional) to role/person anyway.
Do we need a new machinery for this? If a role or person object acquires a new inverse-key relationship, I would find it reasonable to alert the 'mnt-by:' and 'notify:' targets as a matter of course. I don't see the value in defining new attributes just to cover this.
3. The -c flag: This could be incorporated with the irt as well. Or in a even more generalized way have a "whois -c abuse-c 131.130.7.44" returning the most specific inetnum containing an abuse-c, and maybe if useful extending the semantics to other attributes if useful.
I like "whois -c abuse-c 131.130.7.44".
If those three things are incorporated, abuse-c has all the features irt has, except for the creation policy. But this is up to discussion anyway.
I hope I make some sense here
I think so.
lG uk MfG NO8