Path: utzoo!attcan!uunet!husc6!bloom-beacon!tut.cis.ohio-state.edu!triceratops.cis.ohio-state.edu!karl
From: karl@triceratops.cis.ohio-state.edu (Karl Kleinpaste)
Newsgroups: news.admin
Subject: Re: Suggestion for new newsgroup creation rule.
Message-ID: <16833@tut.cis.ohio-state.edu>
Date: 1 Jul 88 19:46:29 GMT
References: <960@ficc.UUCP> <11275@agate.BERKELEY.EDU> <2378@inco.UUCP>
Sender: news@tut.cis.ohio-state.edu
Lines: 54
In-reply-to: mack@inco.uucp's message of 30 Jun 88 20:53:58 GMT

mack@inco.uucp writes:
   It wouldn't necessarily have to be the alt hierarchy. How about a "trial"
   hierarchy? Hey, Backbone! Yeah, you. Will you propagate the trial
   hierarchy?

I'll propagate darn near anything.  It matters very little to me
whether another item lands in my sys file or not.

Now as to whether the whole backbone group would carry it, I dunno.
Sys file complexity matters much more to some than others.  Consider
the site with 160 newsfeeds - it matters to that site's admin a lot.
Your best bet is to write mail to backbone@{purdue,rutgers}.edu and
ask.  When having lunch with the rest of the backbone a week ago
Wednesday, I was a little bit surprised to find that some backbone
folks have really extensive kill files and hence don't read many
discussions in news.{groups,admin} after the 2nd or 3rd day.  There is
a certain point to that; the constant rehash-of-rehash-of-rehash of
complaints and countercomplaints that is typical here makes it rather
pointless for lots of people.  I don't use kill files myself; but I do
`n' through an awful lot at times.  The effect is similar.

[Now, how many people have I angered by saying that there are backbone
admins who don't read Every Single Word of news.{groups,admin}?  Lots,
probably.  Do you think it's justified?  I do.  Consider volume, and
trying to get some real news management done once in a while.]

By the way, I find myself Really Fond of the idea of maximum lifetime
limits within the proposed {new,trial,whatever}.* distribution.
Someone suggested 3 months max; personally, I'd like to pull it back
to 1 or 2.  Also, if it were to exist (I'm not suggesting its creation
tomorrow or anything, mind you), I'd like the trial newsgroups to be
created as trial.full-name-in-regular-namespace, e.g., a group like
trial.soc.widgets as the trial group for soc.widgets.  Then you can
spend part of your time arguing over whether the place where it landed
in trial would be appropriate without the leading string `trial.'  One
hopes that such argument would *stay* there, too, and out of news.groups.

   For maximum flexibility, newsgroup proposers (proposeurs?) could choose
   between requesting a group in the standard (Spaffordized) hierarchy,
   with discussion period and voting, and immediately creating the group
   in the trial hierarchy.

Um, that's liable to cause some troubles, methinks.  Consider the
problems of extremely fast proliferation of newsgroups within the
{new,trial}.* hierarchy.  I could find that somewhat objectionable,
enough that I might stop carrying it.  I'd like at least a certain
amount of solid support for it first.

Just for grins, how many people out there have run into the 512-line
limit on your active file already?  I recompiled 3 or 4 weeks ago for
a 1024-line limit.  I actually hit 512 with a local newgroup
yesterday.

--Karl