[db-wg] Re: abuse-c: proposal
- Previous message (by thread): [db-wg] abuse-c: proposal
- Next message (by thread): [db-wg] Re: abuse-c: proposal
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Niall O'Reilly
niall.oreilly at ucd.ie
Fri Feb 6 16:56:31 CET 2004
Comments in context below ... On 6 Feb 2004, at 14:31, Randy Bush wrote: >> Attribute name >> Name should be distinctive, eg: 'abuse-mailbox'; not 'e-mail' > > since it should be a handle, which allows me to even get to > a <gasp!> phone number, calling it mail-anything seems a > mis-direction. since all the rest are of the form <foo>-c:, > do we have to invent something new? Yes, we do. A number of ppl (MarcoH, dfk), in different ways, and for divers reasons, have made it clear that we need a new attribute to carry as its value (yes, I do mean RHS) the e-mail address where abuse notifications should be sent. Marco explained this particularly well; The following example is an attempt to illustrate his point. Running whois -h whois.ripe.net 137.43.0.0 | \ perl -ne 'next unless /@/; ($addr) = grep /@/, split; print "$addr\n"' gives a bunch of addresses, which will probably ALL be used by your generic cheap-n-cheerful, save-the-world-from-SPAM robot. [ Hmm.. look at those victims! Must try to spread the load even wider! 8-) ] We aim to make it easy for everyone; the following should be nearly enough. whois -h whois.ripe.net 137.43.0.0 | \ awk '/^abuse-mailbox:/ {print $2}' > >> Handle >> Handle is upwardly scalable >> Handle could refer to person, role, or maintainer > > yep > >> Value >> Value is necessary; if not in primary object, then indirectly >> Value in primary object is attractive at low end of scale >> Value is vulnerable to RHS-hijacking, even if not in primary object! > > do not understand. do you mean the value (right hand side) > of the abuse-c: attribute? or are we sending cash? :-) [ I'll happily receive cash. 8-) ] I mean the RHS of the abuse-mailbox: attribute. This attribute MAY be specified in the primary (inet*num or such) object, but for scalability belongs in the object whose handle is on the RHS of a *-c: or mnt-by: attribute. [ Hmm.. *-c, *-by, *-fly 8-) ] As dfk has pointed out, we can get by without abuse-c: if we focus on taking advantage of _existing_ person/role/mntner objects linked by *-c: or mnt-by: and just use abuse-mailbox:. <p loaded-with="irony" remark="In case it's not obvious"> This is the speedy-deployment option, so I'm sure we won't want to use it! </p> IHTH TGIF /NO8
- Previous message (by thread): [db-wg] abuse-c: proposal
- Next message (by thread): [db-wg] Re: abuse-c: proposal
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
[ db-wg Archives ]