At 22:49 -0400 2003-04-22, Ed Agro wrote:
>At 06:16 PM 4/22/03 -0500, you wrote:
>>  > My question is, why does catalist report different numbers of subscribers
>>>  for the same list at the machine name and domain name? I get:
>>>  I notice  that with other lists, the difference can be in the other
>>direction.
>>
>>Well, but are those reports at *the same time*?
>
>Oh, I should have mentioned - the data I reported are for a time when there
>was no subscriber activity. One  of the reports gives a number that matches
>what I get via REV, the other is off. (I just rechecked this.) So the
>question remains.  -  ed

If you bore down further, you'll find that Catalist has different
details on the two lists and on the two servers. Obviously, it
doesn't know that they are names for the same server (and hence the
same list).  Since the one of the domain names is a CNAME to the
other, they are the same server.

My guess is that at one time the server thought of itself as
listserv.acsu.buffalo.edu primarily, but now listserv.buffalo.edu is
its primary identity.  It would be reporting in to Catalist as
listserv.buffalo.edu and updating the descriptions and subscriber
counts for the lists associated with that name, but the descriptions
and subscriber counts for lists associated with
listserv.acsu.buffalo.edu would be left over from the last time it
reported in under that name.

An alternative explanation is that the server once reported in under
both names and was switched to reporting only the primary name.
Also, SUNY Buffalo may have used the two names for different servers
during a server transition.

If you search Catalist for lists with buffalo.edu in the host name,
you will see that many of the lists are shown on both names, but a
few are on only one or the other.  Those would be the lists created
or deleted since the server changed the name under which it reports.
Indeed, if you follow the web archive link of one of the acsu-only
lists, the server that answers doesn't see any archive for that list.

Somebody should probably clean up the duplicate and stale list
entries.  I'm not sure whether this somebody is State University of
New York at Buffalo or L-Soft.

--
Mark R. Williamson