Path: utzoo!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!wuarchive!zaphod.mps.ohio-state.edu!samsung!uunet!inmet!justin
From: justin@inmet.inmet.com
Newsgroups: news.groups
Subject: Re: Results of sci.aquaria vote
Message-ID: <41800012@inmet>
Date: 30 Nov 89 21:35:00 GMT
References: <10413@encore.Encore.COM>
Lines: 52
Nf-ID: #R:encore.Encore.COM:10413:inmet:41800012:000:3017
Nf-From: inmet.inmet.com!justin    Nov 30 16:35:00 1989


/* Written  3:09 am  Nov 24, 1989 by edhew@xenitec.on.ca in inmet:news.groups */
In article <41800010@inmet> justin@inmet.inmet.com writes:
>In other words, I'm not calling an "only-sysadmins-vote" system unfair. I'm
>saying that it is totally, utterly, and completely *pointless*...

How would my "sysadmin vote" be "pointless" considering that I would
have solicited the opinions of my users (being a generous sort of person
who values the opinions of the intelligent individuals I allow to have
access to my system)?  Certainly, in the final analysis, I'm sure that
I weigh my opinion more heavily than that of junior users I just created
accounts for.  Those that participate and contribute currently have
respected comments evaluated, as opposed to ignored.  I doubt that *my*
users would disagree.  It is precisely the "interest" factor that I do
gauge, within the established hierarchy that we appear to have agreed to.
[...]
  Ed. A. Hew, SCO Authorized Technical Trainer, XeniTec Consulting Services
/* End of text from inmet:news.groups */

Given the situation you describe (a sysadmin who really cares about his
newsfeed, and really cares about what the users are going to want), then
using representation has its merits. Unfortunately, I think that you are in
the minority amongst sysadmins. Given the remarkable number of admins who
appear to follow Spaf's list, without bothering to think for themselves
about what's in it; given the number of systems that say, "Oh, alt is just
a gutter, so we won't bother with anything in it," I'm forced to conclude
that there are a *lot* of sysadmins out there who simply don't care as
much as you do...

And even if we did have a perfect world of caring, thoughtful sysadmins,
it still misses the point. Say we have two sites, one single-user PC at
home and a hundred-user mini at a company. Using the simple-minded 
representative scheme people have been talking about here, each would
be equally represented. Now, that's perfectly *fair* (at least, in some
sense), but it means that we are learning little about the actual utility
of the group, if a single reader and a hundred possible readers are
equally represented. (Devil's Advocate: "It might, however, be sufficient
information to determine whether the newsgroup is worthwhile in terms of
cost distribution." I confess, I haven't thought the math through carefully
enough to give a firm opinion here. If it *did* work, it would be based on
very different *sorts* of numbers than we're using today, distribution
percentages instead of number of readers...)

Also, I suspect that we'd cut our sample population down too far. As is,
Usenet "votes" only draw a few hundred people, out of, what?, a hundred
thousand? If the vote was restricted to sysadmins only, we'd have a
"polled" population so small that the noise in the statistics might well
overwhelm any real information we might hope to glean...

					-- Justin du Coeur

"The idea of trial newsgroups looks a little better with each passing week..."


Brought to you by Super Global Mega Corp .com