Xref: utzoo news.groups:8579 news.admin:5335
Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!purdue!haven!mimsy!mangoe
From: mangoe@mimsy.UUCP (Charley Wingate)
Newsgroups: news.groups,news.admin
Subject: No Deletion Rules
Message-ID: <16610@mimsy.UUCP>
Date: 29 Mar 89 12:40:04 GMT
References: <3580@ficc.uu.net>
Reply-To: mangoe@mimsy.umd.edu
Organization: U of Maryland, Dept. of Computer Science, Coll. Pk., MD 20742
Lines: 25

I remain unconvinced that we have any need for these rules, because I don't
see any real reason for these knds of votes in the first place.

There are basically two kinds of deletions: those of active groups, and
those of dead groups.  As far as active groups are concerned, there is
clearly every reason in the world to forbid these kinds of votes; as the
current heat shows, they largely serve as rhetorical tools in content or
personality fights.  Deletions of active groups are purely the prerogative
of sytem adminstrators who do not wish to carry the group, and more
generally, the backbone.

I also think that deletions of dead groups are properly an administrative
function.  The sound reasons for deleting them-- space and overhead-- are
the system admin's purvue, not the readership's.  The bad reason, namespace
size, is in my view not a real problem.  To delete ten or even thirty groups
is not going to make a meaningful dent in the size of the namespace, and
people are still going to be lazy and not bother to try and find out what
groups are appropriate to their posting.

If we need a deletion rule, it should be something like this: when the
propagation of a group, as measured in the arbitron system, gets
sufficiently low for some set number of measurements, it should be dropped
by the backbone.

C. Wingate