The meaning of "whois"

[Dave Crocker <dhc@xxxxxxxxxxxx>]

At 10:57 AM 9/5/2001, Randy Bush wrote:
> > Netsol broke whois a lot of times in the past years by changing the
> > output format and the world has not ended yet.
>
> they did not break whois, they broke (or changed) their application's
> use of the protocol.  naturally, this affected the users of the netsol
> application.


A serious source of confusion appears to be the multivariate use of the 
term "whois".

Within the IETF, I believe we are consistent to mean use of the current, 
IETF standards-track Whois protocol.  Outside the IETF it tends to mean use 
of that protocol AND use of some unspecified set of enhancements (or 
additional protocols).

Both within the IETF and outside of it, we are *IN*consistent in 
referencing the database that is used.  And there are lots of DBs to choose 
from, all in substantial use.  Subsets of the DBs have consistent formats, 
but different content.  Some have consistent semantics, but different 
content.

So a technical discussion probably needs to be explicit about

        1) protocol
        2) port
        3) syntax
        4) semantics
        5) data source / author / owner  (eg, RIPE vs. NetSol.)

Use of the term Whois:43 covers requirements 1 & 2.  I suggest we get into 
the habit of using that notation within this group, since it is what the 
BOF had as its focus and agreed to limit itself to.  Choice of item 5 will, 
of course, tend to specify 3 & 4, so the useful consideration of item 5 is 
probably in terms of "categories of use" such as RIR vs. DNS registry.

We need to work on clarifying items 3-5, in terms of available choices and 
specific selections.

d/

----------
Dave Crocker  <mailto:dcrocker@xxxxxxxxxxxxxxx>
Brandenburg InternetWorking  <http://www.brandenburg.com>
tel +1.408.246.8253;  fax +1.408.273.6464