Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.3 4.3bsd-beta 6/6/85; site mit-eddie.MIT.EDU
Path: utzoo!watmath!clyde!burl!ulysses!allegra!mit-eddie!gds
From: gds@mit-eddie.MIT.EDU (Greg Skinner)
Newsgroups: net.news.group
Subject: Re: New newsgroup proceedures, voting.. (long)
Message-ID: <1232@mit-eddie.MIT.EDU>
Date: Sun, 9-Mar-86 15:43:53 EST
Article-I.D.: mit-eddi.1232
Posted: Sun Mar  9 15:43:53 1986
Date-Received: Wed, 12-Mar-86 01:41:24 EST
References: <5075@alice.uUCp> <3326@sun.uucp> <1195@mit-eddie.MIT.EDU> <1986@hao.UUCP>
Organization: MIT Lusers and Hosers Inc., Cambridge, Ma.
Lines: 180

In article <1986@hao.UUCP>, woods@hao.UUCP (Greg Woods) writes:
> 
>   No reason why Spaf should always have to take all the heat, so I will
> take a stab at responding to this one. Obviously, the person who wrote
> this has no idea what is involved in administrating a backbone site.
> Let me see if I can enlighten you.

Obviously, you have forgotten that not too long ago, I administrated
the hub of netnews for a large computing facility (25 machines).  I was
on the critical path from ihnp4 to most of the other large computing
facilities in Bell Labs.  I think that qualifies me to speak on this
subject (which I have spoken before, in fact in argument with this
same person, who hasn't changed from his last stance, by the way), so
let me proceed.

> 
> > My model of how the USENET voting procedure would run is similar to
> > that of the electoral college.  (short description)
> 
>    Sounds good, but I'll wait until I see a reasonable implementation.
> See my previous article in this discussion. What's a "feed"? Does a long
> distance "feed" get more votes than a local one? What if they send us
> more than we send them? In short, good idea in theory, but an accounting
> nightmare (does that phrase sound familiar to anyone? :-)

Like I said, there are many factors which can go into the weighting.
I agree that long-distance feeds should be weighed more heavily than
local calls.  Volume of news transmitted should also be a factor (if
you bear the brunt more than someone, naturally you should have more
of a say in how the entire net is run).  As I think of more things
I'll list them, but I agree -- it is an accounting nightmare.

> > There are those who seem to have a problem with this scheme.  I wonder
> > why this is so.
> 
>   Simple practicality. Let's take a "close-to-home" example, with my recent
> cut of the "soapbox" groups from the hao system. Suppose the net as a whole,
> including my vote(s), decides that we should keep the soapbox groups.

Before you go on, note that my scheme *in no way* dictates how a host
handles its news locally.  You are free to store whatever articles you
want, and pass any others you want.  The point of all this is to
decide in a democratic manner what groups on a netwide scale will be
kept, added, and so forth.  This way, groups like net.internat will
get removed by consensus, rather than by force.  (Same for net.flame.
I believe under my propsed scheme net.flame would have been removed a
long time before it actually was.)

> I tell my management this, and they say "off the net". Of course, that isn't what
> I would do. What I would probably do is cut the groups anyway, which results
> in an even worse situation that we currently have. Those "downstream" sites
> still get net.philosophy. You are correct that the sites that are willing
> to carry the groups should be allowed to continue to do so even if I drop
> them. But if they do, they shouldn't call the group net.something. They should
> be calling it co.philosophy, because that's what it effectively is there
> (unless someone has started a back door feed for those groups, which
I doubt).

Two points:  (1) You are assuming that a site downstream from you
won't pick up on a group you've dropped, which is unfair to that site.
(Example:  ihnp4 was still carrying net.flame for a while after the
backbone stopped carrying it.  Other sites could have picked it up
from there, or wherever it was going, whichever is cheaper/faster ...)
(2) You seem to think the whole net will vote against the backbone all
the time.  I don't think this is true.  In addition, the backbone
won't all vote the same way (again, net.internat).

> The *FACT* is that the backbone sites WILL decide what net.* groups will
> exist whatever votes are taken. You cannot tell us what we have to pay for.
> And, equally true, I can't tell YOU what NOT to pay for. Any vote taken
> that doesn't have the approval of the backbone sites is a joke. Or at least,
> it isn't a NET group.

Do I sound like I am telling someone how to spend their money?  I
don't believe so.  All I'm saying is, let's do things in an organized,
democratic way.

> > I guess what I am tryng to say is that fair is fair.  What goes on in
> > USENET should not be decided by a select few, it should be decided in
> > a democratic manner.
> 
>   Wrong. It is and will be decided by those who bear the brunt of the cost.
> No amount of voting will reinstate net.flame, net.bizzare or net.religion
> on my machine, and since we pay most of the cost of transporting the news
> into Colorado, we effectively "vote" for all of those sites too, like it
> or not (in case you get the wrong idea, I'm not too thrilled about having
> to cut people off from news they want, but I'm constrained by budgets just
> like any other Gramm-Rudman site).

No one's asked you to put those groups back on your machine (how many
times do I have to say this?).  And since every site that is on the
net bears some percentage of the cost of running the net, every site
should have that percentage of a vote in how the net is run.  Period.
That's democracy.

> > There should be some agreement among those who
> > participate in USENET that yes, you will have to play by the rules, or
> > else you have to get out of the game.
> 
>   I agree, except that what you mean by "the rules" and what I mean by that
> are different. We pay for it. You get what WE pay for (of course, if you
> want to pay for it yourself....)

Well, judging from my map data, I don't get news from you (relief,
because I'm sure I have riled you up enough to cut my feed off :-).
The rules include, but are not limited to, how much a site pays for
news (and for distributing it).  For example, netetiquette.
Supposing someone on a backbone sites constantly violates
netetiquette.  (no spelling flames please)  Does the fact that they
are on a backbone paying for everything give them the right to spew
their excrement all over the network?

> > (I know no such agreement
> > exists now, but in the future, failure to comply with the by-laws of
> > the net should be met with the possible suspension and/or removal of a
> > site from the net.)
> 
>   Oh? And what happens to the sites downstream from us if you decide to
> remove us from the net? Are you SURE that's the way you want to proceed?

I suppose it is better to let those who have the money get away with
all kinds of sh*t, because they are paying for it.  Sounds like a real
good way to run a network.  The net will last such a LONG time too,
because everyone will continue to put up with the crap those who have
the money dish out (but THEY control everything, so we have to do what
they say ...) 


> > I wonder, for those
> > who favor a backbone-run network, do you realize the danger inherent
> > in dictatorship?
> 
>   I sure do. I hear you wanting to tell us how we have to spend our money.
> *THAT* sounds like dictatorship.

This is the last time I'm going to say it.  No one's telling you how
to spend your money.

*sigh*

Somehow, I don't believe I have gotten my point across.  Even if I
have, to some of you, anyone on the backbone who insists that *they*
know what is best for the rest of the network, will continue to do
things in a high-handed manner.  At any rate, let me just say this:  I
do not think it is unreasonable to incorporate voting as a standard
procedure for making netwide decisions, like if a group is to be
created or removed.  I do think it is unreasonable for a few individuals
to make that decision for the rest of the net (where a few individuals
does not necessarily equal the entire complement of the USENET
backbone).  To illustrate a few last examples, when Henry Spencer
refused to pass certain groups, I had no complaint -- in fact, I would
have probably done the same.  However, it is not the same when Gene
Spafford rmgrouped net.bizarre and net.internat.  He clearly took
matters into his own hands.  (For the record, I don't disagree with
why net.bizarre was rmgrouped, I just disagree with the manner in
which it was rmgrouped.)  My voting scheme is set up mainly to
prevent those sorts of actions.  It does not presume to dictate the
actions a site, or group of sites, take locally (like when net.flame
wasn't carried anymore, although I think it was a pretty sneaky way for
the backbone to get their way, sort of like someone picking up their
marbles and going home leaving the other kids with nothing to play
with, I still can't object to it because, collectively or no, it was a
decision carried out individually).  In general, all I'm trying to do
is protect those sites who actually care about doing things in a fair
manner from some twit who, for laughs, sets themself up as a USENET
funder, and proceeds to create and delete groups at whim, because they
hold the purse strings.  (Extreme example, but I think anyone who
understands what I am really trying to say will get my point.)

Sorry for the length, I was pretty steamed when I saw Greg Woods'
response to me -- I hope I was able to communicate my thoughts
adequately.


-- 
It's like a jungle sometimes, it makes me wonder how I keep from goin' under.

Greg Skinner (gregbo)
{decvax!genrad, allegra, gatech, ihnp4}!mit-eddie!gds
gds@eddie.mit.edu