30 day voting: too long?
Chuq Von Rospach
First, a bit of history: the original 30 day voting period was set up
before nntp links were common. If you look at the uunet statistics, 90%
of the messages hit uunet within one day. Two years or so ago it took
about 4 days for the 90% mark to be hit.
Another thing to keep in mind is that originally, that 30 day voting also
included what's now called the discussion phase, so while the time to
distribute a message has gone down, we've actually increased the amount of
time it takes to create a group.
Now, uunet is uunet and isn't completely representative of distribution, but
it's indicative of the changes in speed with which distribution is made.
Stuff that used to get out in four or five days is getting out in one or
two. The kind of distribution delays that made longer wait-times useful have
pretty much gone away. Technically, we don't need long voting times any more.
Socially and organizationally? I don't know. I think 30 days is too long.
I'd like to see it 14 days, but that's probably too short for some groups
and doesn't take into consideration holidays or vacations. So, for the sake
of argument, how about 21 days?
I've asked Greg this question and neither of us can remember a single voting
instance where the results of a vote changed after 14 days. I know that
comp.sys.mac.hardware is in the midst of a runaway -- the voting was 114-9
after 18 hours and is now somewhere around 150-15 after about 36 hours. I
can't see the final results being in question after a week, much less three
weeks.
So, let me ask the following questions:
o Why should we *not* consider reducing the voting to 21 days? What does 30
days give us that three weeks does not?
o Has anyone out there run a vote (or know of one that was run) where the
final result would have been changed if there was a 21 day voting period?
Groups that would not have been created under the new not-yet-proposed
rule that were created (or the that failed but would have succeeded?)
And if we can't find an answer for the latter question, doesn't it sort of
answer the former?
chuq
--
Chuq Von Rospach <+> Editor,OtherRealms <+> Member SFWA/ASFA
ch...@apple.com <+> CI$: 73317,635 <+> AppleLink: CHUQ
[This is myself speaking. No company can contr
Segmentation Fault. Core dumped.
richard welty
*Socially and organizationally? I don't know. I think 30 days is too long.
*I'd like to see it 14 days, but that's probably too short for some groups
*and doesn't take into consideration holidays or vacations. So, for the sake
*of argument, how about 21 days?
*I've asked Greg this question and neither of us can remember a single voting
*instance where the results of a vote changed after 14 days.
two of the votes i've run would not have passed at 14 days, but would
have passed at 21 days.
i agree that 30 is probably unnecessarily long; 21 would be ok;
i'd be opposed to only 14.
richard
--
richard welty 518-387-6346, GE R&D, K1-5C39, Niskayuna, New York
..!crdgw1!lewis.crd.ge.com!welty we...@lewis.crd.ge.com
Officer: Do you know how fast you were going?
Driver: No. The speedometer only goes up to 85
Zaphod Beeblebrox
(in article <34...@apple.Apple.COM>)
|
|I've asked Greg this question and neither of us can remember a single voting
|instance where the results of a vote changed after 14 days.
Er, on the vote I ran for rec.music.dementia, the magic 100+ line was
crossed somewhere around day 20-23. After the first 4-5 days, there
was about a 65-70 vote margin, which then grew somewhat gradually to
the final 110-115 margin. (I don't remember the exact numbers, and
I'm not about to go look up the votes, but these numbers are close.)
B.E.E.
--
Z. Beeblebrox |
(alias B.E.E.) | [ This space for rent. Reply for details. ]
b...@cs.purdue.edu |
PEDRO SAIZAR (614-292-7881)
>
> o Why should we *not* consider reducing the voting to 21 days? What does 30
> days give us that three weeks does not?
I'm going to call for votes for the creation of soc.culture.
latin-america next September 21. If this doesn't invalidate the
voting, I would propose to have a 21-day period and test this
idea. I would REALLY like to have a somewhat shorter voting
period.
So, I will explicitly state a 21 voting period in the call
for votes, bearing in mind that the "rules" are "suggested"
rules in USENET. If there is a problem with this, please, I would
like to hear from any News authority, regulation keeper or
whoever may be concerned.
Pedro
Thanks for listening and tell your friends about the News mix
(adapted from WXMX).
-------------------------------------
Internet: pe...@mps.ohio-state.EDU
Bitnet: pedro@ohstpy
Greg Woods
> I'm going to call for votes for the creation of soc.culture.
>latin-america next September 21. If this doesn't invalidate the
>voting, I would propose to have a 21-day period and test this
>idea. I would REALLY like to have a somewhat shorter voting
>period.
The main reason why a fixed-length voting period is in the guidelines
is so that the vote-taker can't terminate the vote as soon as it has
enough votes to pass. This was done a couple of times in the past and
resulted in flame wars from people who didn't get to register their NO
votes. I don't think it really matters how long the voting period actually
is; the important thing is that when it will end must be announced
IN ADVANCE of accepting any votes, and once announced, the voting
period must last as long as was stated in the intial announcement,
even if it gets enough votes to pass before then, and may not be extended
beyond the initially-announced time.
I don't see anything wrong with shorter voting periods. It takes way
too long to get a group created as it is. Unless there are serious objections
during this discussion, I will amend the guidelines to state that votes
can be taken for any time period from 14-30 days, as long as it is
announced ahead of time how long it will actually be.
--Greg
Chuq Von Rospach
>is so that the vote-taker can't terminate the vote as soon as it has
>enough votes to pass.
Definitely. The thing that bothers me, though, is that we're holding back
the non-controversial groups because of the generally rare special case.
I'd rather see things move relatively quickly, with some way of slowing down
the process if there isnt a clear agreement.
My current thought: cut the voting period to 21 days (two weeks is too
short, at least until we get enough experience at 21 days to be able to
judge it better). If, however, the voting is exceptionally lopsided, allow
it to be terminated early and the newgroup to happen.
By exceptionally lopsided, I mean voting the following:
o After 14 days, if the group qualifies under the standard rules *and* the
affermative votes are 80% of the total votes, the group is considered
to have passed.
Rationalization for this is that if a group is up 150-10 after seven days,
it's *very* unlikely that the no-voters are going to change the final
outcome....
>I don't think it really matters how long the voting period actually
>is; the important thing is that when it will end must be announced
>IN ADVANCE of accepting any votes
Agree, mostly.
First, I think it's wrong to allow someone to just arbitrarily stop a vote
and declare a winner. On the other hand, if the voting is a runaway, though,
there's a point where waiting any longer is silly (my analogy to a couple of
people have been softball games. In many leagues, if there's a 10 run lead
halfway through, they all call it and go drink beer instead).
I also think that it *does* matter how long the voting period is. I feel two
weeks is a minimum. Assume a person reads USENET once a week. To guarantee
he sees the voting request, you basically have to have it around for about
two weeks thanks to propogation and timing. (if they can't get through
news.announce.newgroups weekly, I'd say they aren't motivated enough about
the subject to be worth worrying over). Also, less than two weeks and you
start running into people's vacations, holidays, etc.
The three week voting with a 14 day cuttoff for runaways is, I think, a
reasonable compromise between speed and making sure people have a chance ot
get their preferences noted. I'd even argue for a 7 day cutoff for an
exceptional runaway, but I'm not convinced that's really long enough for
people to react to -- it'd be possible for a well-organized interest group
to shove the ballot box full and close voting before other folks could react.
Greg Woods
>
>Definitely. The thing that bothers me, though, is that we're holding back
>the non-controversial groups because of the generally rare special case.
Yes, but that's better than turning a non-controversial case into a
controversial one by terminating the vote early. I really think the only way
to avoid flame wars is to have the termination date announced IN ADVANCE,
and then stick to that termination date NO MATTER WHAT the preliminary
results are. Yes, it may slow down the creation of an obviously favored
group, but anything else is GUARANTEED to generate ridiculous flame wars.
Personally, I really don't think that the need for ANY new newsgroup is
so critical that it can't wait another week to be created. It isn't
worth the potential controversy just to save one lousy week. (At least, that's
how *I* view it; obviously Chuq doesn't agree. What does everyone else think?)
>Rationalization for this is that if a group is up 150-10 after seven days,
>it's *very* unlikely that the no-voters are going to change the final
>outcome....
I agree, but they will still flame like hell if they aren't given a chance
to vote. Remember that one of the main purposes of having a vote is to shut
up the losing side. Allowing votes to be terminated early will make it
harder to accomplish this, and really doesn't gain enough to be worth it.
>I also think that it *does* matter how long the voting period is. I feel two
>weeks is a minimum. Assume a person reads USENET once a week. To guarantee
Well, OK. I should have been more clear. Yes, I think we have to have some
reasonable minimum, but I think 30 days is too long. We also need some
reasonable maximum, so that a given issue doesn't drag on forever. I think
14-30 days is reasonable. But I'd be willing to go with 21-30 days if most
of you think that's better.
--Greg
Chuq Von Rospach
>(At least, that's
>how *I* view it; obviously Chuq doesn't agree. What does everyone else think?)
I could live with it either way, actually. I just feel like sometimes we're
flogging a horse that is not only dead, but rapidly rotting. There are times
when it feels like we've nominated God to the Supreme Court, but still have
to wait 30 days to see if someone is going to challenge His qualifications...
> Well, OK. I should have been more clear. Yes, I think we have to have some
>reasonable minimum, but I think 30 days is too long. We also need some
>reasonable maximum, so that a given issue doesn't drag on forever.
Actually, I think we ought to have *a* voting time. Otherwise we get into a
confusing range of voting dates that differ at the whim of the vote
coordinator (and, frankly, almost all of them are going to choose the
shorted possible vote time anyway, just to get it over with. The only reason
to choose a longer vote is to try to squeak in as many votes as possible for
a really marginal group, and that doesn't seem like a good reason to have a
sliding scale for voting lengths....)
Brad Templeton
done have been obvious within 2 days. But, yes, it's important it
be announced in advance just how long it is. 3 days would be enough
for me if the group champion felt that the current 100 amount would be
reached by then. And in spite of what some self-important people think,
it doesn't matter in a survey (unlike a vote) that any given individual's
response is counted, it only matters that the selection factor is sufficiently
random.
--
Brad Templeton, Looking Glass Software Ltd. -- Waterloo, Ontario 519/884-7473
PEDRO SAIZAR (614-292-7881)
> [discussion deleted]
Heck. What a conversation we have here! Can I jump in now?
I think that the smallest set of "rules" for group creation would be
very convenient for those of us that only once in a lifetime will
undertake such a thing. I think that a free choice of a voting period
between 15 and 30 days, according to the time and enthusiasm of the
vote taker, would be the best solution.
Now, some time ago, I made a suggestion that almost nobody commented
or it was diverted to a completely different subject. My idea was to
pay more attention, not to group creation, but to group inactivity.
An inactive group could be set "on probation" for sometime and, if
still inactive, elliminate it. If a group receives, say, 50 messages a
year, probably it's not worth keeping it and there will surely be
another group dealing with similar stuff. In that way, their existence
is not decided by occasional voters, but actual readers and writers.
Is it possible to elliminate a group from the News once it was created?
Ah, and I think that percentages are a better indicator than differences.
One more thing: if I got them right, the statistics tell us that groups
are read by thousands of people, in a net with half a million users. So,
what in the world does a vote of 150-12 mean? Non-sense. Less than 0.1%
is actually voting? Or I'm misunderstanding something?
Good night, Oh my 136,953 readers out there!
Pedro Saizar
Internet: pe...@mps.ohio-state.EDU
Bitnet: pedro@ohstpy
------------------------------------
William B. Thacker
>In article <34...@apple.Apple.COM> ch...@Apple.COM (Chuq Von Rospach) writes:
>>
>>Definitely. The thing that bothers me, though, is that we're holding back
>>the non-controversial groups because of the generally rare special case.
>
> Yes, but that's better than turning a non-controversial case into a
>controversial one by terminating the vote early. I really think the only way
>to avoid flame wars is to have the termination date announced IN ADVANCE,
I think that, so long as the rules are clearly stated in the guidelines,
there will be little question; on par with what we get anyway, "My
no vote wasn't counted !"
>Personally, I really don't think that the need for ANY new newsgroup is
>so critical that it can't wait another week to be created.
This is true. But I suspect that many, like myself, wait eagerly
for the 30-day period to get over with when an interesting, new
group is proposed. We're only talking convenience here.
>>Rationalization for this is that if a group is up 150-10 after seven days,
>>it's *very* unlikely that the no-voters are going to change the final
>>outcome....
>
> I agree, but they will still flame like hell if they aren't given a chance
>to vote. Remember that one of the main purposes of having a vote is to shut
>up the losing side. Allowing votes to be terminated early will make it
>harder to accomplish this, and really doesn't gain enough to be worth it.
The voting for sci.military showed that, over the duration of the vote,
the yes-to-no ration remained essentially constant. I do suggest that
an important consideration is that the discussion does continue during
the vote. A late-breaking issue might point out a change in
the character of the group from what was originally perceived, causing
a change in the yes-to-no ratio and even the reversal of some votes. I
personally think this would be a minor effect, and can't think of any
live examples.
I once suggested (I think) that the vote-taker could terminate a vote
whenever she had achieved *twice* the 100-vote margin for success, if
she wanted to go for early group creation. I concur, though, with Chuq's
suggestion of 14 days as a bare minumum. Could we then, say:
"A vote lasts up to 30 days. If, at any time after the first 14 days,
there are at least 200 more Yes votes than No votes, they vote-taker
may opt to declare the vote ended and post the results. If the
vote goes the full 30 days, only 100 more Yesses than Nos are required
for group creation."
I feel that the extra 100 vote margin should be more than enough to
quench any flames; and I believe several recent votes could have met
this criterion.
- - - - - - - - valuable coupon - - - - - - - clip and save - - - - - - - -
Bill Thacker w...@cbnews.att.com
"C" combines the power of assembly language with the flexibility of
assembly language.
Karl Lehenbauer
a bad thing.
21 days sounds OK.
Someone'll bitch if their vote doesn't quite pass by the end of a shortened
voting period.
There's always alt for those who can't wait.
--
-- uunet!ficc!karl "Have you debugged your wolf today?"
Todd Day
~I want to start a bit of a meta-discussion: is a 30 day voting period too long?
yes
~o Has anyone out there run a vote (or know of one that was run) where the
~ final result would have been changed if there was a 21 day voting period?
I ran the vote for comp.dsp. After one week, it was running about 115-8.
After two weeks, it was ~200-14. The final vote after 30 days was 265-18.
As you can see, the ratio of no votes to yes votes stayed pretty constant
(~7%) throughout the voting period. I could have shut it off after two
weeks, or even a week, and the results wouldn't have changed.
What is really ridiculous is that we only allow five days to fix any voting
errors. I think we should make it 28 days, with voting closing after 21
and a whole week to fix any voting errors. Also, if there are 200 more yes
votes than no votes *at any time*, voting stops and the group is created.
--
Todd Day | to...@ivucsb.sba.ca.us | ivucsb!to...@anise.acc.com
"Just give me a killer sound system and the babes will follow."
The Cat in the Hat
news.groups's own b...@cs.purdue.edu (Zaphod Beeblebrox) said:
-Said ch...@Apple.COM (Chuq Von Rospach):
-(in article <34...@apple.Apple.COM>)
-|
-|I've asked Greg this question and neither of us can remember a single voting
-|instance where the results of a vote changed after 14 days.
-
-Er, on the vote I ran for rec.music.dementia, the magic 100+ line was
-crossed somewhere around day 20-23. After the first 4-5 days, there
-was about a 65-70 vote margin, which then grew somewhat gradually to
-the final 110-115 margin. (I don't remember the exact numbers, and
-I'm not about to go look up the votes, but these numbers are close.)
All the more reason for a 14 day voting period... [ 1/2 :-) ]
I've never minded having the 30 day vote period, but then there's never
really been a group that I've been so excited about that I couldn't wait
the 30 days for. I do like Chuq's landslide rule though...
if you don't care, does
it matter how long it takes?
--
David Bedno, Systems Administrator, The Santa Cruz Operation, Inc.
Email: dav...@sco.COM / ..!{uunet,sun,ucbvax!ucscc,gorn}!sco!davidbe
Phone: 408-425-7222 x5123 Disclaimer: Speaking from SCO but not for SCO.
"Your reality is lies and balderdash, and I'm glad to
say that I have no grasp of it." - Baron Munchausen
Brad Templeton
to be created. (Although in some cases, like sci.physics.fusion, the group
got created way too late. -- and if you say, 'that shouldn't have been created
anyway,' are you arguing that we want to create short-duration groups after
their lifespan is waning?)
The problem is that currently it takes 2 months to create a group,
which is just too much hassle. I don't want 2 months of work to create
A group -- 2 months of discussion, vote announcements, tallies and argument.
Erland Sommarskog
>I want to start a bit of a meta-discussion: is a 30 day voting period too
>end a vote when the result is clear.>
Greg and Chuq have already discussed this a lot, let me just add my
few pennies.
Greg is reluctant to accept prematurely ended votes to avoid flame
wars. I don't think this is a problem, *if the rules are set from
the beginning*. If the guidelines say that a vote should last for
21 days, but the vote-taker has a right to terminate it if at any
time he has, say, 200 more YES votes than NO votes. The vote must
last for at least seven days, though. As long as this is part of
the rules, and not left for the vote taker's arbritarity, there'll
be nothing to feed the flames.
I think it's more risky with giving the vote taker the option to
choose how long the period should be. Then you can flame him for
taking a too short/long period.
So, I think that the voting period should be cut down, but the
guidlines should clearly state 21 days, and nothing else. Of
course combined with a likewise clearly state cut-off rule.
Note that the 200 above was just an example. I feel that a lower
absolute majority combined with some percentage would be better.
For instance, 120 in absolute numbers and 90 in percent.
One could argue that some people would dissatisfied not getting
the chance to vote because they were on holiday the entire
week the vote lasted. In this case the result is very clear,
and their votes would not affect the result, so the votes would
be more of a symbolic nature. And if they are that hot on newsgroups
they shouldn't go on holiday without a Usenet connection in the
first place.
>o Has anyone out there run a vote (or know of one that was run) where the
> final result would have been changed if there was a 21 day voting period?
> Groups that would not have been created under the new not-yet-proposed
> rule that were created (or the that failed but would have succeeded?)
The case is of course a group that barely passes. The vote I took on
rec.music.cd was a thriller. The majority was steadily increasing,
but not until day 28 or 29 I finally passed the magic 100. (The
group finally passed with 144-43.)
The conclusion of this is that with a shorter period a slightly
less number of groups will pass. I see nothing wrong in that.
--
Erland Sommarskog - ENEA Data, Stockholm - som...@enea.se
Chuq Von Rospach
>Greg is reluctant to accept prematurely ended votes to avoid flame
>wars. I don't think this is a problem, *if the rules are set from
>the beginning*.
I tend to agree with this, but I also see Greg's point and think we can take
this a step at a time. No sense moving too fast and creating new problems by
not thinking things through.
>Note that the 200 above was just an example.
I will, again, recommend *strongly* that we use percentages instead of
hard-coded numbers. There should be a minimal number (100) to make sure
there is a general interest for a group, but using a number like 200 only
guarantees that down the road when we're getting 200 votes all the time
we'll just have to argue this all again and set new limits. One of the
functions of an ever-growing net is that a voting limit that's reasonable
now is going to be easy to surpass in six months or a year -- the number of
votes being cast, on average, is continuing to grow and we shoudl plan for
that.
--
Chuq Von Rospach <+> Editor,OtherRealms <+> Member SFWA/ASFA
ch...@apple.com <+> CI$: 73317,635 <+> [This is myself speaking. I am not Appl
Segmentation Fault. Core dumped.
Brad Templeton
validitity of a survey, there is a general issue worth pointing out.
It's *good* too have a fair degree of vagueness and discretion in teh
guidelines.
Half the argument in news.groups seems to be about claims of technical
violation of the guidelines by survey-takers! People argue about no votes
that missed by 2 hours, as though the guidelines were some cast in stone set
of rules.
Now if the guidelines said, "the survey taker can end the survey any time
he or she wishes, 14 +/- 2 days after starting it" we could not have argument
over exact following of the guidelines, because by definition it would be
hard for the survey taker to be in violation!
People forget that these are *not* rules, and they argue like they were. So
let's make the guidelines deliberately an explicitly inexact. Never say
14 days, say 14 +/- 1.
Bob Sloane
> Make it 7-30 days. While some surveys take longer, all the ones I have
> done have been obvious within 2 days. But, yes, it's important it
> be announced in advance just how long it is. 3 days would be enough
> for me if the group champion felt that the current 100 amount would be
> reached by then. And in spite of what some self-important people think,
> it doesn't matter in a survey (unlike a vote) that any given individual's
> response is counted, it only matters that the selection factor is sufficiently
> random.
The problem with making the vote time too short is that proponents of the
group might very well organize before the vote was announced and get their
votes in early. Imagine a vote proposal from a site where it takes 2 days for
the call for votes to reach uunet. I would probably be 3 days before a
majority of sites got the proposal, and another 3 for the votes to get back to
the vote taker. The vote taker could email people likely to vote yes in
advance, saying that the vote would begin on such and such a time. I suspect
that one could get just about any group passed if allowed to make the voting
period short enough and pick the site to handle the voting.
One thing I have never really understood is why you think of the newsgroup
creation process as a survey. If this process were really a survey, shouldn't
we select a random sample of usenetters and send them a request asking how
they feel about the newsgroup creation? I think of the group creation votes
as a referendum on the creation of the group. My dictionary says:
Referendum n. 1. The submission of a proposed public measure or actual statute
to a direct popular vote.
Survey n. 1. A detailed inspection or investigation. 2. A comprehensive view.
[ommiting the parts that deal with surveying land.]
Which definition do you think fits better?
--
USmail: Bob Sloane, University of Kansas Computer Center, Lawrence, KS, 66045
E-mail: slo...@kuhub.cc.ukans.edu, slo...@ukanvax.bitnet, AT&T: (913)864-0444
"The scientific theory I like best is that the rings of Saturn are composed
entirely of lost airline luggage." -- Mark Russell
Stanier A
?year, probably it's not worth keeping it
Why? What costs are incurred by keeping a newsgroup with few
messages?
--
Alan M Stanier * University of Essex, Colchester CO4 3SQ, England
email al...@essex.ac.uk |
voice +44 206-872153 | Should you understand this article, please contact
fax +44 206-860585 | me. I shall gladly explain it until you don't.
Brad Templeton
>The problem with making the vote time too short is that proponents of the
>group might very well organize before the vote was announced and get their
>votes in early.
If somebody wants to try this sort of trick to bias their survey, they can
do it now. If they declared a 4 day survey for
comp.culture.celts.druids.asterix, and it was an amazing success, but was
posted from a backwater site, they would be asking for trouble. Ok, I
would be content to make a minimum of 7 days, to avoid potential flame wars
about fraud.
>
>One thing I have never really understood is why you think of the newsgroup
>creation process as a survey. If this process were really a survey, shouldn't
>we select a random sample of usenetters and send them a request asking how
>they feel about the newsgroup creation?
It is a survey not because of how it's done, but because of what it is.
You are correct, a good survey is done as you describe. We have surveys
conducted sort of like votes, which adds to the confusion.
It is not a vote because it's non-binding. I can issue a newgroup message
at any time. Each individual site can decide to honour it as they see
fit. The survey result merely helps each admin decide that it's a good
thing. By consensus, but not law, admins usually decide that the groups
with good enough surveys will get passed around in general. I can also
ignore any survey -- there are many groups I don't carry on this machine.
Or I, or any group of admins could decide on their own criteria. We might
decide to use a 200 margin instead of 100, or to use a ratio.
Does that sound like a vote to you? Votes are binding, draw usually from
a set of registered voters, and sometimes have security.
Your dictionary defined a good survey, that's what was confusing!
Greg Woods
>The problem is that currently it takes 2 months to create a group,
Not to nitpick (OK, I'm nitpicking :-) but it currently takes exactly 7
weeks (49 days) to create a group, assuming one follows the guidelines. This
is more like 1.5 months than 2. But I still agree it's too long.
I also tend to agree with Todd Day when he said that 5 days is too short for
vote verification.
--Greg
Greg Woods
>If the guidelines say that a vote should last for
>21 days, but the vote-taker has a right to terminate it if at any
>time he has, say, 200 more YES votes than NO votes. The vote must
>last for at least seven days, though. As long as this is part of
>the rules, and not left for the vote taker's arbritarity, there'll
>be nothing to feed the flames.
Sure there will. Say someone starts a vote. After 5 days (I was away
for a long weekend) I send in a NO vote. After 7 days the vote is
published with a 200+ majority, and my vote isn't counted. I now
suspect fraud since the voter's site and mine are both on the Internet
but he didn't get my vote (or so he claims) for 2 days. It is this kind
of argument that I wish to avoid. It is true that we could possibly
have this kind of argument now at the end of a 30 day period, but it is
much less likely since most people have voted by then. I just don't see
any advantage to giving the vote-taker the power to terminate the vote
earlier than the date announced when the vote started. The need for
even overwhelmingly favored groups just isn't that critical in most
cases. In a case of a major topical group, the current rules don't
cover that at all (e.g. sci.physics.fusion). In that event we ought to
either work out a method of creating temporary groups or create the
group under alt. I just don't like the idea of allowing the vote-taker
to influence the result in any way (other than his right to cast a
vote, of course. When I ran the vote for news.announce.newgroups, I
didn't even do that. I thought it would do more harm drawing flames
than good).
>I think it's more risky with giving the vote taker the option to
>choose how long the period should be. Then you can flame him for
>taking a too short/long period.
Chuq convinced me of this as well. The only thing that will minimize (note
I didn't say eliminate) flames is use the same voting period for ALL votes,
no matter what. No argument in that case. But I still think 30 days is too
long.
--Greg
slo...@kuhub.cc.ukans.edu
> In article <11...@kuhub.cc.ukans.edu> SLO...@kuhub.cc.ukans.edu (Bob Sloane) writes:
>>One thing I have never really understood is why you think of the newsgroup
>>creation process as a survey. If this process were really a survey, shouldn't
>>we select a random sample of usenetters and send them a request asking how
>>they feel about the newsgroup creation?
>
> It is a survey not because of how it's done, but because of what it is.
> You are correct, a good survey is done as you describe. We have surveys
> conducted sort of like votes, which adds to the confusion.
>
> It is not a vote because it's non-binding.
Interesting. We have non-binding referendums on the ballot here. The last one
I remember was concerning a fee on our water bill. They are mostly held to
allow the city commission to find out what people want the commission to do.
Sounds a lot like what you describe as a "survey."
> I can issue a newgroup message
> at any time. Each individual site can decide to honour it as they see
> fit. The survey result merely helps each admin decide that it's a good
> thing. By consensus, but not law, admins usually decide that the groups
> with good enough surveys will get passed around in general. I can also
> ignore any survey -- there are many groups I don't carry on this machine.
Exactly my point. The purpose of a referendum is to find out what people
think about something, at least around here. On the issue I mentioned above,
most people voted that they didn't want the fee. That didn't mean that the
city commission was required by law to get rid of the fee, it just meant that
people thought they should. No legal mandate, just a consensus. I don't
know, but I suspect that is where the name "REFERendum" came from, ie,
refering some issue to the "public" by means of a vote. Sort of like asking
for advise.
> Or I, or any group of admins could decide on their own criteria. We might
> decide to use a 200 margin instead of 100, or to use a ratio.
>
> Does that sound like a vote to you? Votes are binding, draw usually from
> a set of registered voters, and sometimes have security.
Why are votes binding? If we have a referendum on issuing bonds in our
county, and it passes, do you really beleive that the county is legally
required to issue the bonds, even if the project is canceled for other
reasons? You seen to think of a vote as somehow "electing" something.
Referendums don't work that way at all. In many cases, they are EXACTLY like
newsgroup votes. Non-binding, informational only.
> Your dictionary defined a good survey, that's what was confusing!
Why think of newsgroup votes as bad surveys when you can think of them as good
referendums?
Brad Templeton
We can't do much to avoid that. We should work on getting something
that avoids flames from people who (are supposed to) know how things work.
We can eliminate endless technicality flames by explicitly relaxing the
guidelines. The harder it is to differ from the guidelines by accident,
the harder it is for people to (semi-justifiably) make flames about it.
To eliminate the pointless fraud flames (Fraud flames are the last resort
of the incompetent) the only real answer is to have a program on an
independent machine receive and count the answers. I and others have
suggested this dozens of times by now, and everybody seems to think it's
a good idea. If I had a central enough machine to do it, I would do it.
When is somebody going to volunteer? Your machine *can* handle the load,
believe me, because news.groups will be so much smaller in return!
Werner Uhrig
> In article <2...@enea.se> som...@enea.se (Erland Sommarskog) writes:
>
> Sure there will. ..and my vote isn't counted. I now suspect fraud
well, yes, I can see it. some people always would complain ...
> The only thing that will minimize (note I didn't say eliminate) flames
> is use the same voting period for ALL votes, no matter what.
the ONLY thing? thought of ALL alternatives? Minimize?
doesn't leave much room for other opinions, does it now...
yes, for SOME people thjs would undoubtedly be true; However, I
think MOST people on this here net are tolerant people who see
this net as a cooperative friendly effort of communicating useful
information, and the last thing they want to see is that the
pathetic bureaucratic needs of SOME people should ruin any chance
for flexible rules, interpreted in the spirit of trying to make
nearly everyone happy.
In the particular case at hand, I think that taking turns at collecting
votes nearly assures an impartial vote-count; and by sending all
votes to either one (trusted) central site (or to the vote-gatherer
AND one central vote-archiving site) MANY netters might feel that
their net-liberty and rights are reasonably safe.
I dare say, that MOST netters I know, quietly enjoy the good parts
of the net, and when asked, they'd like it most if those screaming
bone-heads with fixed solutions for everything would just go away...
...including me, sometimes ... :-)
--
-----------> PREFERED RETURN-ADDRESS FOLLOWS <--------------
(ARPA) wer...@rascal.ics.utexas.edu (Internet: 128.83.144.1)
(UUCP) ..!utastro!werner or ..!uunet!rascal.ics.utexas.edu!werner
Greg Woods
>Newcomers will flame no matter how we set things up, I'm fairly sure.
True. But not all flame wars involve newcomers!
>We can eliminate endless technicality flames by explicitly relaxing the
>guidelines.
No, we can't. If the guidelines are changed to read "30 days +/- 1"
instead of "30 days", people will still flame if it goes for 32 days.
The only thing that will lower the amount of "technicality flames" is if
people stop to think about whether the INTENT of the guidelines has been
followed rather than the LETTER. For example, if someone waits for 32 days
to post vote results to allow time for those last couple of YES votes to
trickle in and push the vote over the top, that is clearly what the limit
is designed to prevent. If instead the extra days are because the vote-taker
was on vacation when the 30th day came around, that is clearly only a violation
of the letter rather than the intent. It is violations of the INTENT that we
really need to be concerned about. I doubt if Brad would argue with me about
that; the question is, how do we get the technicality flamers to see this?
Answer: we probably can't.
>the only real answer is to have a program on an
>independent machine receive and count the answers.
This is a reasonable proposal that ought to be considered. However, it is
not without some problems that would have to be addressed. For one thing, it
will create new types of flames from people who send in improperly formatted
votes that don't get counted. Secondly, I don't believe such a thing as an
"independent" machine exists. Everyone, including site administrators, has
opinions. And even if not, can we really trust someone who doesn't care about
these issues to do the work required to ensure that votes all run to completion?
Lastly, it does complicate the voting process. No longer would you just
be able to reply to the call for votes article, you now have to know a specific
address to send votes to. I'm not saying these obstacles are necessarily
insurmountable; it's just that I would need to see these issues addressed
satisfactorily before I would favor implementation of such a proposal.
>everybody seems to think it's a good idea
This is surely an exaggeration, as this article proves. I am not the only
one to voice some of these objections.
>Your machine *can* handle the load,
Machine load isn't the issue. The issues are 1) getting the net to accept
this plan, which requires addressing at least some of the points brought up
in the previous paragraph; and 2) Having the time and energy to write the
required software. If it weren't for these I would happily volunteer MY machine
as the collection point.
>believe me, because news.groups will be so much smaller in return!
That is debatable. The flamers will always find SOMETHING to flame about, no
matter WHAT we do!
I would like to take this opportunity to thank Brad for taking the time
to put forth a real proposal for what we could do better rather than just the
usual bitching about the current setup. All he will probably get for his
trouble is a lot of flames.
--Greg
Pedro Saizar (614-292-7881)
> In article <43...@ohstpy.mps.ohio-state.edu> PE...@ohstpy.mps.ohio-state.edu (PEDRO SAIZAR (614-292-7881)) writes:
>
> Why? What costs are incurred by keeping a newsgroup with few messages?
>
So, what's the big deal with creating new newsgroups? Why such a
discussion about voting and making sure if it is worth or not creating
them? Are there any other reason besides the costs involved?
Cheers, (my favorite show)
Pedro
Amanda Walker
(Pedro Saizar (614-292-7881)) writes:
> So, what's the big deal with creating new newsgroups? Why such a
> discussion about voting and making sure if it is worth or not creating
> them? Are there any other reason besides the costs involved?
In a word, "history." Several years ago (that is to say, in the Deep Mists
Of Time as far as Usenet is concerned), newsgroups were created because
someone thought they should be, and were removed for the same reason. This
worked while the net was small, and some people claim it still works for
the "alt" newsgroups.
Then Usenet exploded. Cheap UNIX boxes, cheap fast modems, NNTP, and so
on. With more people, and more idiots, controls (or the appearance of
them, anyway) were instituted. This helped keep a few people from
disrupting the net for everyone else with "newgroup/rmgroup wars" and
endless discussions about what exactly a newsgroup should be called.
For a more detailed history, try Spaf's regular "the history of Usenet"
posting.
--
Amanda Walker
ama...@intercon.com
richard welty
*In article <16...@looking.on.ca> br...@looking.on.ca (Brad Templeton) writes:
* If instead the extra days are because the vote-taker
*was on vacation when the 30th day came around, that is clearly only a violation
*of the letter rather than the intent.
why would this ever happen? the one time i was on vacation when
a vote ended, i just checked the received-from lines in the header
to find out when the votes physically arrived in my in box.
richard
--
richard welty 518-387-6346, GE R&D, K1-5C39, Niskayuna, New York
..!crdgw1!lewis.crd.ge.com!welty we...@lewis.crd.ge.com
T. William Wells
: of the incompetent) the only real answer is to have a program on an
: independent machine receive and count the answers. I and others have
: suggested this dozens of times by now, and everybody seems to think it's
: a good idea. If I had a central enough machine to do it, I would do it.
: When is somebody going to volunteer? Your machine *can* handle the load,
: believe me, because news.groups will be so much smaller in return!
OK. I volunteer. But I don't think that it will make any difference.
Why?
Because I doubt that the net will accept a single vote counter.
Why?
Because every vote taker is going to want to do it his way. But if
I'm counting the votes, I'll do it my way. I may be very easy about
what I permit, but I will have some rules. And rules==flames, at
least on Usenet. :-(
And:
Because one vote taking machine isn't enough. If Joe Random can
commit fraud, so the argument will go, so can I. Never mind that
every instance I'm aware of where fraud has been charged has been a
case of sour grapes. (I'm of the opinion that a charge of fraud
should get the charger hauled into court for libel unless there is
real evidence, and not the "my vote didn't make it" whines, for it.)
---
Bill { uunet | novavax | ankh | sunvice } !twwells!bill
bi...@twwells.com
David C Lawrence
brad> of the incompetent) the only real answer is to have a program on an
brad> independent machine receive and count the answers. [...]
brad> When is somebody going to volunteer?
In <1989Sep20....@twwells.com> bi...@twwells.com (T. William Wells):
bill> OK. I volunteer. But I don't think that it will make any difference.
I have volunteered similarly in the past and am doing so again now.
Dave
--
(setq mail '("ta...@pawl.rpi.edu" "ta...@itsgw.rpi.edu" "ta...@rpitsmts.bitnet"))
Raymond Dunn
>was on vacation when the 30th day came around, that is clearly only a violation
>of the letter rather than the intent. It is violations of the INTENT that we
>really need to be concerned about. I doubt if Brad would argue with me about
>that; the question is, how do we get the technicality flamers to see this?
>Answer: we probably can't.
This seems an easy one to solve by giving the vote-taker the right to make
some arbitrary decision about "late" votes, then the intent of the law is
that the vote-taker can bias late votes any way his whims direct him (:-).
Let the rules state that the voting period be a *minimum* number of days,
after which the vote can be stopped *at any time* by the vote-taker, up to
some maximum, lest the period be infinite.
So, specifically, I propose that the voting period be a minimum of 21 days
and a maximum of 30. The actual time being arbitrarily decided by the
vote-taker, and not pre-announced.
So, to paraphrase a certain Larry Laffer ("Save early, save often"):
"Vote early, vote, eh..., once".
--
Ray Dunn. | UUCP: r...@philmt.philips.ca
Philips Electronics Ltd. | ..!{uunet|philapd|philabs}!philmtl!ray
600 Dr Frederik Philips Blvd | TEL : (514) 744-8200 Ext : 2347 (Phonemail)
St Laurent. Quebec. H4M 2S9 | FAX : (514) 744-6455 TLX : 05-824090
Dave Sill
>So, specifically, I propose that the voting period be a minimum of 21 days
>and a maximum of 30. The actual time being arbitrarily decided by the
>vote-taker, and not pre-announced.
This seems like excellent compromise. 21 days should be sufficient
for virtually all potential voters to receive the vote call and for
their vote to reach the ballot box. It should be clear after 21 days
whether or not the group will pass. If it's close after 21 days, or
the vote-taker wants to hold out for a last-minute rush of YES votes,
then he can wait up to 9 days longer. Everone gets a chance to vote,
and landslide votes can be ended sooner.
--
Dave Sill (ds...@relay.nswc.navy.mil)
Greg Woods
>In article <7...@philmtl.philips.ca> r...@philmtl.philips.ca (Raymond Dunn) writes:
>If it's close after 21 days, or
>the vote-taker wants to hold out for a last-minute rush of YES votes,
>then he can wait up to 9 days longer.
I am strongly opposed to this. After 21 days no doubt we would see another
call for those last few votes. I think if a group proposal is this marginal
it doesn't NEED to pass. This is just needless fuel for flames.
--Greg
richard welty
=In article <1...@ark1.nswc.navy.mil> ds...@relay.nswc.navy.mil (Dave Sill) writes:
=>In article <7...@philmtl.philips.ca> r...@philmtl.philips.ca (Raymond Dunn) writes:
=>If it's close after 21 days, or
=>the vote-taker wants to hold out for a last-minute rush of YES votes,
=>then he can wait up to 9 days longer.
= I am strongly opposed to this. After 21 days no doubt we would see another
=call for those last few votes. I think if a group proposal is this marginal
=it doesn't NEED to pass. This is just needless fuel for flames.
i agree. in particular, for marginal groups, this allows the vote taker
to select a cutoff point so that it falls between the yes vote that
puts the group over the top and the no vote that pushes it back down.
this would not be a good situation.
Stanier A
?Think about organizing a filing cabinet. Would you leave a bunch
?empty, labled folders hanging around ? Not after the second or third
?time you went looking for something that was misfiled or just hard to
?find. Who wants to keep looking in empty folders ?
?
That is a lousy analogy. Newsgroups signal the fact they have unread
articles in them, whereas folders do not.
In any case, we are not talking about empty newsgroups, but ones with
little traffic. That traffic will not go away if the newsgroup does, but
will be added to some other newsgroup - through which readers will then
have to wade to find the articles that interest them. Deleting little-used
newsgroups will make Usenet harder, not easier, to use.
--
Alan M Stanier * University of Essex, Colchester CO4 3SQ, England
email al...@essex.ac.uk |
voice +44 206-872153 |
fax +44 206-860585 | If in doubt, do.
Chuq Von Rospach
>?empty, labled folders hanging around ?
> That is a lousy analogy. Newsgroups signal the fact they have unread
>articles in them, whereas folders do not.
> In any case, we are not talking about empty newsgroups, but ones with
>little traffic.
No, it isn't a lousy analogy, because you're thinking in terms of reading
newsgroups. The real problem is one of posting to the correct newsgroup.
Think of this situation: you are a manager for a large computer company. You
hire a temporary secretary for two weeks to take that three inch stack of
papers and file it into your filing cabinet. You tell him to read each paper
and file it in the folder that is most appropriate (but that he can't make
new folders).
Now, is this person going to be more likely to find the proper folders if
there are 400 folders or 600 folders? And, to tie this *back* to reading,
if you then want to go find one of those papers he filed, what are the
chances that you and he are going to agree on the name of the folder? Are
you more likely to agree if there are 600 folders? or 400?
Realistically, whether the filer (in USENET jargon, the poster) and the
reader find the same folder depends entirely on how well the folder names
have been defined (the USENET name space), how unambiguous the material is
and how well the filer can think. USENET has the advantage of cross-posting
to reduce some of the ambiguity/lookup problem, but I feel that using that
as a 'solution' is red-herring: the filer, for instance, could simply file a
Xerox of each letter in each file and we'd have 100% lookup success, but a
total failure as a filing system. Optimally, we should never have a need to
cross-post. Reality isn't optimal, but the % of cross-posting should be
taken as an indication that the group isn't well named or well defined (or
is unnecessary, except as a symbolic link to other newsgroups).
The analogy is actually pretty good. The problem with the USENET name space
is the same as a filing cabinet -- how can Joe Random Novice User figure out
which folder to stick his message in while (1) minimizing the number of
copies he has to make to cover the options, and (2) make sure that the
people who are looking for the message are looking in the same place he's
putting it.
That's the problem with too many group names, and groups that aren't used.
They make it hard for the novice to find the *proper* place for postings
that aren't somewhat ambiguous. That's why we have naming space wars: to try
to make it easier for the new kids, not to protect the experts. The experts
have already figured it out (I hope!).
Gil Pilz@Eng@Banyan
>In article <43...@ohstpy.mps.ohio-state.edu> PE...@ohstpy.mps.ohio-state.edu
>(PEDRO SAIZAR (614-292-7881)) writes:
>?If a group receives, say, 50 messages a
>?year, probably it's not worth keeping it
> Why? What costs are incurred by keeping a newsgroup with few
>messages?
Cognitive costs, increased search time for every user on the net
(especially new users) . . . another possible newsgroup to check if
your looking for the "optimal" group to discuss something or ask
questions in.
Think about organizing a filing cabinet. Would you leave a bunch
empty, labled folders hanging around ? Not after the second or third
time you went looking for something that was misfiled or just hard to
find. Who wants to keep looking in empty folders ?
The net's namespace is not a garbage receptacle for unused crap, it's
a *tool* for locating the specific information _you_ are interested in
out of the huge mass of information flowing around. It's not a very
good tool but it's the one we've got. Ideally you would like the
majority of people to be able to trace down the appropriate
heirarchies to find the groups they're interested in. Deadwood groups
make this more difficult by increasing the search space.
Gilbert W. Pilz Jr. g...@banyan.com
Banyan Systems Inc. (617) 898-1196
115 Flanders Road
Westboro, MA 01581
Peter da Silva
> I could live with it either way, actually. I just feel like sometimes we're
> flogging a horse that is not only dead, but rapidly rotting. There are times
> when it feels like we've nominated God to the Supreme Court, but still have
> to wait 30 days to see if someone is going to challenge His qualifications...
I'm afraid I take strong objection to this analogy. There is no place for
God on the Supreme Court of the United States, and I for one would join
in many others in challenging any such nomination.
"The Government of the United States is in no way founded on the Christian
Religion" -- George Washington.
--
Peter da Silva, *NIX support guy @ Ferranti International Controls Corporation.
Biz: pe...@ficc.uu.net, +1 713 274 5180. Fun: pe...@sugar.hackercorp.com. `-_-'
"That is not the Usenet tradition, but it's a solidly-entrenched U
delusion now." -- br...@ucsd.Edu (Brian Kantor)
Peter da Silva
> final result would have been changed if there was a 21 day voting period?
Me. comp.unix.i386 just barely passed under the current rules. And it was
proposed for the best of reasons: too much volume in comp.unix.xenix and
comp.unix.microport.
William B. Thacker
>>?Think about organizing a filing cabinet. Would you leave a bunch
>>?empty, labled folders hanging around ?
>
>> That is a lousy analogy. Newsgroups signal the fact they have unread
>>articles in them, whereas folders do not.
>
>newsgroups. The real problem is one of posting to the correct newsgroup.
I disagree, Chuq. When I file something in a folder, it's because I
want to go back and retrieve it at a later time. When I post something
to Usenet, I never intend to see it again, only the followups, and
resultant email.
Of the groups on Usenet, I only subscribe to 25-30. All other are
transparent to me, until I need them. Now, let's say I want to
post something. In a recent example, I wanted to get a recommendation for
new tires for my car. Now, I knew rec.auto.* existed, so I already knew
where to go; but I could have view'd my .newsrc file, searching for
strings such as /tire/ and /car/ and /auto/, or just thumbing through the
rec and misc heirarchy's. Either way, I come up with "rec.auto" as a likely
candidate. I subscribe to the group, and skim the headers for likely
titles; I see that there are many requests for tire info there, but none
of the sort I need. At this point, I could have done a (c)atchup to clear
out all old articles and posted my request; instead, I sent mail to one
of the more knowledgeable posters to the group, whose articles on tires
had caught my eye.
[I will add, though, that it would be nice to have a better method to
search for the appropriate newsgroup. Perhaps each site could keep,
online, a version of Spaf's List of Active Groups, which has lots of
keywords in it for searching on. Almost as good, it would be *awfully*
nice if that list, as well as the list of moderators and mailing lists,
didn't expire so quickly; it's only available 50% of the time ! When
Spaf's list is available, one can check it to be sure one's posting is
appropriate to the group in question]
>Realistically, whether the filer (in USENET jargon, the poster) and the
>reader find the same folder depends entirely on how well the folder names
>have been defined (the USENET name space), how unambiguous the material is
>and how well the filer can think.
Quite agreed, here; but I submit that that's not enough. We need a way for
users to search based on keywords. There's simply no way that the
group name can be specific enough to be sufficient for the task. In
other words, we need an index to the Usenet namespace, just as a
complicated filing system, such as the Yellow Pages, might have an index.
>Optimally, we should never have a need to
>cross-post.
I disagree. There are plenty of times when it's appropriate; calls for
votes being a splendid example. Crossposting should be minimal, of course.
But consider stuff like "Announcement; Benefit Concert to Save the Whales";
crossposting to rec.music.something, alt.activism, and talk.animal-rights
would be appropriate. Oh, and misc.headlines, because *everything* is
crossposted there 8-)
- - - - - - - - valuable coupon - - - - - - - clip and save - - - - - - - -
Bill Thacker w...@cbnews.att.com
"C" combines the power of assembly language with the flexibility of
assembly language.
jeff daiell
>
> There is no place for
> God on the Supreme Court of the United States, and I for one would join
> in many others in challenging any such nomination.
>
> "The Government of the United States is in no way founded on the Christian
> Religion" -- George Washington.
First, the only Constitutional qualification I can think of for the
Federal Supreme Court is citizenship -- and I'd have to read it
again to make sure even *that's* in there. But even if it is, if
God is willing to be naturalized, there would be objection
except on the grounds of his/her ideology -- we shouldn't vote
no on a nominee because of his/her religion.
Second, Peter, you're assuming that God is a Christian.
Third, I believe that quotation originated with John Adams. However,
I will give you a Washington quote to support my first argument:
"Our Constitution ... gives to bigotry no sanction."
Jeff Daiell
Followups to alt.flame.churchandstate.quotations.
Erland Sommarskog
my site though, we seem to have got tired on news, and have got
anything for while. Probably nothing gets out. Probably you're
reading this when it's a month old. Why I am posting this anyway?)
Greg Woods (wo...@handies.UCAR.EDU) writes:
>I said:
>>If the guidelines say that a vote should last for
>>21 days, but the vote-taker has a right to terminate it if at any
>>time he has, say, 200 more YES votes than NO votes. The vote must
>>last for at least seven days, though. As long as this is part of
>>the rules, and not left for the vote taker's arbritarity, there'll
>>be nothing to feed the flames.
>
> Sure there will. Say someone starts a vote. After 5 days (I was away
>for a long weekend) I send in a NO vote. After 7 days the vote is
>published with a 200+ majority, and my vote isn't counted. I now
>suspect fraud since the voter's site and mine are both on the Internet
>but he didn't get my vote (or so he claims) for 2 days.
OK, so there may be flames, but will they be well-founded? Are we
free from flames today? Are we free from people who refuse to
realize that the probability for that lost NO votes would change
the outcome is below 1E-10? At the time for the misc.headlines.unitex
we were not. The only way to get entirely free from flames would
to have that central voting machine, and besides that would be flamed
too.
>When I ran the vote for news.announce.newgroups, I
>didn't even do that. I thought it would do more harm drawing flames
>than good).
Of course you couldn't. It wasn't in the guidelines at the time.
> Chuq convinced me of this as well. The only thing that will minimize (note
>I didn't say eliminate) flames is use the same voting period for ALL votes,
>no matter what. No argument in that case. But I still think 30 days is too
>long.
Agreed. Make it 14 or 21. But make it a constant limit. And the more
I think of it, I think 14 is perfectly OK. With 14 a cut-off rule,
wouldn't probably be worth the effort. (And with 14 a few more groups
would fail.)
--
Erland Sommarskog - ENEA Data, Stockholm - som...@enea.se
Gil Pilz@Eng@Banyan
>Quite agreed, here; but I submit that that's not enough. We need a way for
>users to search based on keywords. There's simply no way that the
>group name can be specific enough to be sufficient for the task. In
>other words, we need an index to the Usenet namespace, just as a
>complicated filing system, such as the Yellow Pages, might have an index.
The *point* though, is that _until_ we have such an index (and I agree
that one is necessary) it would behove us not to make things more
difficult for new users by junking up our "filing system" with a lot
of trash newsgroups.
This whole sub-thread started by someone asking "what's the big deal
with creating new or carrying dead newsgroups, it's just a handful of
inodes and a couple K of disk, right ?". WRONG ! Even perfectly named
newsgroups (of which there are no such thing) increase the size of the
search space you alluded to in your article. An unused newsgroup
increases the size of this search space without carrying any traffic
to justify this "cost".
-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-
Gilbert W. Pilz Jr. g...@banyan.com
Banyan Systems Inc. (617) 898-1196
115 Flanders Road
Westboro, MA 01581
-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-
Peter da Silva
about Spaf's List of Active Groups:
> nice if that list, as well as the list of moderators and mailing lists,
> didn't expire so quickly; it's only available 50% of the time !
I semi-automatically grab that list (and the other lists) whenever it comes
around. It would be nice if he'd put an Archive-name: header in it so I could
grab it completely automatically (I have a program that grabs such marked
messages in the groups I tell it to). As it is I have to tell news to go get
it.