Xref: utzoo news.groups:8406 news.admin:5258
Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!ames!decwrl!labrea!csli!cphoenix
From: cphoenix@csli.STANFORD.EDU (Chris Phoenix)
Newsgroups: news.groups,news.admin
Subject: Re: Proposed OFFICIAL Newsgroup Creation/Deletion Guidelines
Message-ID: <8281@csli.STANFORD.EDU>
Date: 25 Mar 89 08:52:17 GMT
References: <1634@ncar.ucar.edu> <37740@bbn.COM> <27806@apple.Apple.COM>
Sender: cphoenix@csli.Stanford.EDU (Chris Phoenix)
Reply-To: cphoenix@csli.stanford.edu (Chris Phoenix)
Organization: Center for the Study of Language and Information, Stanford U.
Lines: 81

In article <27806@apple.Apple.COM> chuq@Apple.COM (Chuq Von Rospach) writes:
>Make the size of the required margin dependent on the volume. A group that
>is rarely/never used should be easier to get rid of than a group that has a
>high volume. And I disagree that it should just be "keep" votes. Put the
>burden of proof on the people who want to delete a group by making the
>default to keep it. Therefore, the vote should be some percentage of votes.

Wait a minute.  Do we EVER want to delete a group that has a high
volume?  The only reason I can think of for doing this is if it badly
violated net policy, in which case it shouldn't be a votable issue.
I agree strongly that the default should be to keep it.  I can't see
any normal situation in which we would want to delete a group with 
more than, say, 50 readers and 4 postings/month.  These numbers can be
derived from arbitron, can't they?
As I see it, deletion of a newsgroup should just be a cleanup task.
A useful newsgroup should never be deleted (except for violations).
As a user, I define 'useful' to be one that people are using.
By the time a group should be deleted, the only person who would care
enough to call for a vote would be an administrator who wasn't reading
it.  This seems to be a bad way to start off.

>Hmm. A thought just occurred to me. The "100 more yes than no votes" has
>always been somewhat controversial because the absolute number is relatively
>easy to overcome. Thinking about it, a better way is to define succeed/fail
>in terms of percentages. How about this:
>
>o A group creation vote succeeds if 67% of the vote is for creation.
>
>o A group deletion vote succeeds if 67% of the vote is for deletion.

If it wasn't clear above, I'm opposed to voting on group deletion.

>o In both cases, if the number of spoiled or proven missing ballots exceeds
>  2% of the total vote, or if enough spoiled/missing votes are found that
>  would have changed the result of the vote, the vote must be re-taken
>  by an agreed-upon neutral third party.

Given the vagaries of e-mail, this seems a bit harsh.  Far more than 2%
of my off-system letters get bounced.  I'd be very surprised if many 
votes would pass this test.

>o Any vote with less than 150 total votes will be considered failed for 
>  lack of interest.
>
>This puts the bias against creating groups and against deleting existing
>ones. A group that cares about it's survival should be able to roust enough
>votes to stay above the 33% limit. 

One thing that I haven't seen yet--who would vote for deleting a newsgroup?
Obviously, most of the people reading it wouldn't.  People with a vendetta
would, but I would hate to have a process that would be swayed by such 
a group.  The only other people I can think of are administrators.  As 
you are suggesting it, you would basically have to have 100-150 administrators
vote against a group to get it deleted.  This could quickly stir up 
resentment among them, for having to go through this every time they want
a group deleted, and against them, for trying so hard to delete a group 
they don't really care about.
As far as I can see, this would be forcing lots of people in positions of
power to either vote as an unthinking block, or take sides on something
they don't care that much about, whenever they want to get something 
done.  For some reason, I just don't like the idea.

>And it sets up the success to be relative
>to the total interest in the group rather than trying to set some arbitrary
>number (with a low-end limit set in to keep ennui from taking hold).

IMHO, there are many better ways of determing interest than a vote.

[suggestions for moderated groups deleted]
I don't know enough about how moderated groups work to have any firm basis
for opinion.  A few ideas, though:
1) The only reason for deleting a group, moderated or unmoderated, should
be total lack of interest or bad violations of rules.  In neither case is
a vote appropriate.
2) Making a group unmoderated, or changing moderators, is a tough one.
Seems like the decision should be up to the group alone.  Is there any
way of telling who has been subscribed longer than X weeks, and accepting
only their votes on such an issue?

Chris Phoenix
cphoenix@csli.stanford.edu