[ppml] Last Call for Comment: Policy Proposal 2003-5

[william at elan.net]

There are several companies publishing data as general whois. Here is an 
example:

NetRange:   207.99.0.0 - 207.99.127.255
CIDR:       207.99.0.0/17
NetName:    NAC-NETBLK01
NetHandle:  NET-207-99-0-0-1
Parent:     NET-207-0-0-0-0
NetType:    Direct Allocation
NameServer: NS1.NAC.NET
NameServer: NS2.NAC.NET
Comment:    ADDRESSES WITHIN THIS BLOCK ARE NON-PORTABLE
Comment:
Comment:    * Reassignment information for this network is available
Comment:    * at whois.nac.net 43
RegDate:    1996-04-23
Updated:    2001-08-22

So I do not find anything wrong with CW using their rwsl whois server 
(in fact my recursive whois service will automaticly sublookup in CW rpsl 
whois). The point is that protocol used for publishing data should be 
understood by ARIN and by end-users. The wording "ARIN approved" as has 
been introduced by Mark Kosters seems quite good choice to me to say what 
protocol is to be used. 

What is lucking is possibly process of approving new protocol, but here 
I'm not sure it should really be a policy, in my opionion approval of new 
ready protocol should be done on recomendation of AC (or ARIN staff) by 
BoT and  should be matter of adding new protocol to some list to be 
published on arin website. 

On Tue, 22 Apr 2003, Larry J. Blunk wrote:

> 
> > On Tue, 22 Apr 2003 Michael.Dillon at radianz.com wrote:
> > 
> > > >I was given an augument against making it generic since ARIN only
> > > >allows for ip reassignment information to be of the form of a SWIP or
> > > >RWhois server. If ARIN allows for another form of distributed lookup
> > > >service such as using a LDAP based service (like referral LDAP) or XML 
> > > based 
> > > >service (like IRIS), then it should be made more general at that point.
> > > 
> > > I strenuously disagree.
> > > 
> > > ARIN policies should not contain irrelevant details. In this case, SWIP 
> > > and Rwhois are irrelevant details.
> > > 
> > > The root purpose of this policy is to specify that organizations who have 
> > > received IP address space must keep track of where they are used (maintain 
> > > an IP address directory) and must publish the directory of IP address 
> > > usage. They can choose to either publish the directory by submitting 
> > > regular updates to ARIN for inclusion in a central directory or they can 
> > > publish the directory themselves by maintaining a publicly accessible 
> > > server connected to the Internet 24/7.
> > 
> > whois is the accepted format...whether you swip your data and have ARIN 
> > add it to their whois server or you run your own, it's the only protocol 
> > currently approved.  whois clients are fairly ubiquitous.  I don't have to 
> > know anything about the protocol other than "it's whois" to use a whois 
> > client.
> 
>    C&W and ARIN made a presentation at the Las Vegas ARIN meeting last year
> about C&W using their RPSL-based whois server to report address
> usage (through RPSL "inetnum" objects).  The presentation can be found at
> http://www.arin.net/library/minutes/ARIN_IX/Reporting_Utilization.pdf
> I'm not sure of the status of this project (Tanya? Michael?) or whether
> ARIN plans to allow such reporting outside of C&W.   While the C&W
> server uses the whois "protocol" (a TCP pipe on port 43), the RPSL
> format is considerably different than the output format of RWhois.
> 
>    I'd be careful about substituting the term "whois" for "RWhois" as
> it might imply the usage of RPSL whois servers (in addition to RWhois
> servers) is considered acceptable by ARIN.  I think there should be
> a more formal decision on whether RPSL whois servers are
> acceptable before generalizing the policy to include any "whois"
> service (and, by extension, LDAP/XML/etc. services).
> 
> > If you publish your reassignment data in an LDAP server, how do I access 
> > it?  How do I find out how to access it?
> > 
> > If someone else chooses to "publish" their reassignment data in some
> > obscure CGI deep in their web site and forces anyone who wants to look up
> > an IP to jump through many hoops to get at the data, what then?
> > 
> > If you want to leave the method of reassignment data publication open, 
> > then you're going to have to define a minimum set of guidelines that the 
> > method of publication must meet to qualify.
> >  
> 
>    I concur with this.   Perhaps the policy could refer to some other
> document or policy which contains the guidelines for acceptable
> publication protocols and formats?
> 
> 
>  -Larry Blunk
>   Merit