ln63fkn@sdcc7.UUCP (Paul van de Graaf) (11/02/85)
Just as there is a protocol for starting a newsgroup, there should be a
protocol for ending one. I don't recall a precedence for removing an
active newsgroup until the recent demise of net.bizarre. I'm sure all
know more than a few feathers have been ruffled on both sides of this
issue. Granted: net.bizarre and net.internat were rogue newsgroups, but
a little diplomacy might have helped ease tensions.
Given that removal of an active group has little precedence, one can hardly
expect that newsgroup members would bother reading net.news.group to find
out if such action would take place. I started reading net.news.group when
I saw an off-handed remark in net.bizarre talking about "what was going on
behind my back". After that I got the impression that net.bizarre was on
psuedo probation, and I started sending some of the more blatant misusers
mail to gently remind them to cool it. The volume and quality of postings
improved somewhat, but then came discussions about hanging toilet paper,
simulating cat behavior, and the real kicker - a HUGE prime number posted
not once, but twice! Soon after, net.bizarre was history.
Now we have people complaining they weren't consulted. Not much can be
done about this now, but in the future I hope people wouldn't have this
argument. I have nothing against removing groups per se, but I have some
thoughts about the procedure of removing a group.
Proposed Protocol:
1.) Discuss the problem with the newsgroup in net.news.group.
Establish the reasons for removing it.
Sufficient grounds might be:
high volume, poor signal to noise ratio, commercial content, etc.
2.) Someone like Spaf decides the problem warrants action.
Posts a probation notice in the newsgroup which:
Says why the group is in trouble.
Tells what needs to be done to keep the group.
Encourages members to police themselves and read net.announce.newusers.
Directs discussion to net.news.group.
This message should repeat 1 or 2 times a week to catch new readers and
assure no-one can say he/she missed it.
3.) If after one month (maybe less) the group does not shape up, kill it!
Otherwise the group stays, but remains on probation until the powers
that be decide to lift it.
If anyone wants to try this out on net.sources.mac, go ahead! I especially
dislike posts of binary images. At least with REAL sources, if I can't use
the software in question, I can learn something by reading it.
Paul van de Graaf sdcsvax!sdcc7!ln63fkn U. C. San Diegomjl@ritcv.UUCP (Mike Lutz) (11/05/85)
This proposal (probationary periods for groups that are getting out of
hand) seems to be on the right track. At least it addresses an area
for which there is no well-formed policy, but for which one is
desperately needed: the protocol for newsgroup deletion. I'd like to
add a couple of related proposals, but I want to make certain my
position is clear.
1. There is no doubt that the backbone sites (and many of the
ribs, like us) are being swamped by the sheer volume of news.
2. Most of the volume is generated in a few mega-flamage groups,
some of which, unfortunately, seemed like a neat idea when the net
was younger and smaller. I personally see no reason why we have to
continue to live with the mistakes of our youth.
3. The current newsgroup creation rules are inadequate to address
the problems of volume, yet inhibit the evolution of the net by
making it nigh on impossible to legally go off on a new tangent,
no matter how useful, beneficial, or any other motherhood quality.
So what to do? It seems to me that I've seen 3 distinct outcomes to
the creation of a newgroup, each of which warrants a different
response.
1. The group has a bunch of submissions initially, but eventually
it turns out to be too narrow or too dull. We need some "sunset" rules
to clean up the corpses of these dead groups. I consider this an annoying
but benign problem (like fallen arches).
2. The group has a consistent, moderate traffic volume. "Moderate"
is a subjective term, but that's life folks. Indeed, moderate may
differ from group to group. In any event, such groups are the ones
we should encourage, as they provide valuable service to a wide audience
at a reasonable (acceptable?) cost.
3. The group evolves into a haven for flamers and those who should
be in Hyde Park. Such groups are cancerous, and are the ones threatening
the net's survival. Perhaps we should first try "chemotherapy", like
probation, to see whether we can bring the group under control. If this
doesn't work, we need radical surgery.
The surgery I propose is a new subcategory of net.* called net.sb
(for soapbox groups). Groups considered cancerous would be moved there,
using the alias mechanism. What is more, *no* site would be under any
moral, ethical, or any other constraint to carry such groups. If there
is enough interest in the "topic," the sites interested in it could
still pass the stuff around in any way they liked.
Ok, who decides when a group is dead? What is moderate traffic,
valuable service, acceptable cost? When does a group become
cancerous? Well, as an SA on a (small) rib, I'd like a say, of course.
However, I think the backbone sites should have a bigger say, and I'd
be willing to give up my vote on the assumption that the backbone SA's,
as a group, are rational, reasonable persons. So far, I have no reason
to believe otherwise. Something like a 2/3 majority of the voting
sites would invoke one of these mechanisms.
To keep this from being too secret, there must be a forum for
discussing impending changes. Nothing ticks off the populace like an
unexpected change in the environment. Also, once a decision has been
arrived at by the sites having votes, I think that the result of the
vote (just for/against/abstain) should be published. In the case of
net.sb groups, those willing to keep supporting the groups might want
to advertise that fact.
Comments welcome. Flames are detected by my infallible expert mail
reading system, and will be dropped on the floor (after automatically
replying with a viscious, cutting remark like "So's your old man."
--
Mike Lutz Rochester Institute of Technology, Rochester NY
UUCP: {allegra,seismo}!rochester!ritcv!mjl
CSNET: mjl%rit@csnet-relay.ARPArandy@ranhome.UUCP (Randys Account) (11/05/85)
Paul van de Graaf says, > Just as there is a protocol for starting a newsgroup, there should be a > protocol for ending one. ... > [discussion of demise of net.bizzare] > > Proposed Protocol: > 1.) Discuss the problem with the newsgroup in net.news.group. ... > > 2.) Someone like Spaf decides the problem warrants action. > Posts a probation notice in the newsgroup which: > Says why the group is in trouble. > Tells what needs to be done to keep the group. > Encourages members to police themselves and read net.announce.newusers. > Directs discussion to net.news.group. > > > 3.) If after one month (maybe less) the group does not shape up, kill it! > Otherwise the group stays, but remains on probation until the powers > that be decide to lift it. There has been some discussion lately about establishing/reviewing procedures for removing groups. There does seem to be "demonstrated need" to somehow reduce the amount of "noise" on the net. This would seem to be a good start at a procedure for removing groups. Perhaps removing some of the high volume/least technical groups would be a good place to start trimming. If net.[religion|philosophy|politics] were removed, and net.flame changed to a general net.discussion.of.opinions or something like that, there would be less need to look elsewhere for groups to cut. I would like to say more, but I would rather be brief. -- cbosgd|----------!fortune\ ihnp4 | allegra| \!ranhome!randy dual | hplabs | / topaz |-!pyramid/ (Randy Horton)
lear@topaz.RUTGERS.EDU (eliot lear) (11/09/85)
I would like to say that I agree with Mike Lutz's article
<8998@ritcv.UUCP> in principle. I believe a milder change might be in
order first. If a group is being considered for removal, it would
seem to me that a message should be sent to BOTH net.news.group and
that group saying so. Maybe the best thing to do would be to ask for
votes (including reasons why the group should exist when that is in
question) as to whether there is still enough interest in the group.
If the group doesn't pass the parameters needed to create a new group
then it should be aliased to a group like net.sb.
[The rest of this article applies to creation of new groups as well as
those being considered for removal.]
As far as who votes... I don't care but 2 things MUST be taken into
consideration:
1) Is the group constructive? Is it of any practical / intellectual
value or is it more net.garbage?
2) Is there an audience among the net.community for such a group? I
would love to see a group like mod.economics but I cannot imagine
such a group attracting a large audience. (Maybe I'm wrong about
that too.)
The problem with only polling SAs of backbone sites is that
they cannot show interest in EVERY group on the net. Maybe the SAs of
all sites should locally poll their users and report the info to their
nearest backbone SA who can then vote from that information. (I guess
that is like the Senator/constituent method.) Of course, the problem
here is that the burden falls upon too many people to take a survey
for a group like net.cooking.soups.... In any event, the "Powers that
be" (ie Spaf and company) could use these figures to guide (but not
necessarily govern) them in creating/removing groups. All I am really
trying to say (in an admittedly verbose manner) is that the views of
the masses must be taken into account. Using this method, of course,
would eliminate suprise rmgroups too.
Comments?
Flames > /dev/null.
eliot
--
Should the opinions expressed above be those of someone else besides the
author.. Well.. it ain't my fault.
[lear@topaz.rutgers.edu]
[{allegra,seismo}!topaz!lear]