Xref: utzoo news.groups:10788 news.admin:6235
Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!ncar!woods
From: woods@ncar.ucar.edu (Greg Woods)
Newsgroups: news.groups,news.admin
Subject: Re: tmp.* hierarchy (was: moderated "newsgroups" group)
Message-ID: <3677@ncar.ucar.edu>
Date: 11 Jul 89 21:45:13 GMT
References: <1528@stl.stc.co.uk> <3486@ncar.ucar.edu> <282@ecijmm.UUCP> <8670@cadnetix.COM> <3620@ncar.ucar.edu> <292@ecijmm.UUCP>
Reply-To: woods@handies.UCAR.EDU (Greg Woods)
Organization: Scientific Computing Division/NCAR, Boulder CO
Lines: 95

In article <292@ecijmm.UUCP> jmm@ecijmm.UUCP (John Macdonald) writes:
>There is no difficulty in removing a group - it is a simple control message.

  Ah, if only that were the case... It is a "simple control message" that
must be explicitly acted upon by every site administrator on the net.
Not as simple as it looks.

>When part of the official charter of any tmp group is an explicit time
>limit, then there is not much cause for anyone to object when a rmgroup is
>done at the appropriate time.

  Oh, yeah? I'll bet you 10:1 odds that if the group is still active when
the "agreed-upon" time for expiration comes along, there will probably be
more flames posted than the entire volume of the group up to that point if
anyone attempts to remove the group. Were you around for the net.wobegon 
debacle? Admittedly in that example there was never any agreement to remove
the group ahead of time, but I'm willing to bet it won't make any difference.
People will point to current activity and claim (rightly so) that the group
is still needed.

>The main problem may be seeing 200 different
>rmgroup messages all bouncing around the entire network on X-day as 200
>system administrators get a chance to be backboner-for-a-day.

   As long as all the messages request the same action, this is really
not a big problem.

>That's sort of a cross between the two different possibilities I
>originally suggested (1) urgently needed groups which are "certain"
>to eventually get a permanent home and which are disrupting discussion
>elsewhere, and (2) known temporary topical discussions which are certain
>to never need a permanent home (but for which it may be difficult to come
>up with a good time limit a priori).

  This is where we disagree; I don't think we NEED newsgroups for type (2)
groups. For one thing how do we know the difference between (1) and (2)?
That is what voting is supposed to determine! The proper way to deal with
type (2) groups is to let them get voted into existence, but then have a 
way to rmgroup inactive groups from the regular heirarchy. I think much of
the opposition to the creation of new groups would go away if we could really
count on getting rid of inactive groups.

> }      or just to get around the normal time it takes to go through the
> }"official" group creation procedure.
>
>I don't mind this in exceptional cases, but it should not be the rule.

   Again, how do we define what is "exceptional"? Remember the flame wars
over sci.physics.fusion? They would have been MUCH worse if we didn't
have an alt.fusion to post things to. Now that it looks like there wasn't
really any cold fusion, we would certainly look like fools for defining
this as "exceptional" :-)

>I still see two topical categories - those that *will* fade, and those that
>will not but are urgently needed.

  But how do we know which is which? If it's really going to fade that
quickly, what's wrong with discussing it in an existing group? If it isn't
going to fade that quickly, readers of the related group in which the postings
are appearing will be the first to vote YES for creating a permanent home
for the topic.

> }>>to flames and bad feelings if a group is removed when some people don't
> }>>agree that it has outlived its charter.  
> }
> }   This is easily fixed, by setting a date ahead of time when the group will
> }be deleted.
>This works fine for type (1) groups - either they fulfilled their promise
>and a regular group has been established, or they have not and thereby
>have failed.  However, it is not so easy with type (2).

   The real problem is that there is no way to determine ahead of time which
is type (1) and which is type (2). If we made a distinction whereby one type
has an enforced time limit and another does not, can't you already see the
flame wars over which classification a given topic should get? This might 
itself defeat the purpose of having tmp groups at all.

>I suggested four months for groups that are expected to become permanent.
>This allows some time for the tmp group to operate and be observed before
>a vote is forced.  If there is a demonstrable lack of volume after a month
>and a half, then that later might fail, where an earlier vote would have
>possibly succeeded.

   OK, just to show that I'm not just blindly disagreeing with everything you
say, I can see the point here. In fact sci.physics.fusion is a good example
of this.

>How about an amplification of Rusty Carruth's suggestion.  I would suggest
>that we allow a special fast vote for tmp groups.  It could take place
>without discussion before the call to vote, and require 50 more yes than no
>votes in ten days, or 75 more yes than no votes earlier to cut it even shorter.

  This might be worth considering.

--Greg