jeffd@ficc.uu.net (jeff daiell) (11/29/89)
All of the proposals for new voting schemes -- and they seem
to be getting hairier by the moment -- ignore one fact:
the status quo works pretty well. An occasional abuse,
like con.aquaria, does happen, but I'd hate to see the
current nature of the net abandoned in an overly-zealous
reach to stamp out all possible problems.
A cure can be worse than the problem being addressed.
Jeff Daiell
Elise, are you still out there?
--
"I'm just a soul whose intentions are good ...
Oh, Lord, please don't let me be misunderstood!"
-- The Animalswbt@cbnews.ATT.COM (William B. Thacker) (11/30/89)
In article <7139@ficc.uu.net> jeffd@ficc.uu.net (jeff daiell) writes: >All of the proposals for new voting schemes -- and they seem >to be getting hairier by the moment -- ignore one fact: >the status quo works pretty well. An occasional abuse, >like con.aquaria, does happen, but I'd hate to see the >current nature of the net abandoned in an overly-zealous >reach to stamp out all possible problems. Hear ! Hear ! Is anyone else tempted to draw an analogy to the War on Drugs ? "These are flames. These are the Usenet guidelines on flames. Any questions ?" Let's consider the actual cost of the sci.aquaria debate: 1) Bandwidth. For the month and a half involved, the volume in news.groups averaged well over 50 articles a day; some days, over 100. Most of this was redundant, and much of it was flaming. 2) Personal. The honor and motives of numerous posters has been publically tainted, either by word or deed. I daresay that most of them would rather have been having a root canal than put up with the harassment. 3) Net.structure. We now have alt.aquaria (some places), sci.aquaria (some places), rec.pets.fish (some places), and a vote is in progress for rec.aquaria, which will pass and be everywhere. And I wouldn't be surprised to find comp.birds.cockatoos out there somewhere. Was it worth it ? What was gained ? - - - - - - - - valuable coupon - - - - - - - clip and save - - - - - - - - Bill Thacker AT&T Network Systems - Columbus wbt@cbnews.att.com Black Holes are where God is dividing by zero
alien@cpoint.UUCP (Alien Wells) (12/01/89)
In article <7139@ficc.uu.net> jeffd@ficc.uu.net (jeff daiell) writes: > >All of the proposals for new voting schemes -- and they seem >to be getting hairier by the moment -- ignore one fact: >the status quo works pretty well. An occasional abuse, >like con.aquaria, does happen, but I'd hate to see the >current nature of the net abandoned in an overly-zealous >reach to stamp out all possible problems. Oh really, why don't you check out the current proposal for sci.groupware? Or how about the name controversy in the *.pagan, *.newage, *.nature-worship, *.... proposal going on? Usenet voting has a serious problem. It can't deal with multiple names. If concensus isn't reached (and it often can't be), it is the whim of the proposer to pick the name that will be voted on. If it passes, a better name has no chance. -- --------| Rest assured that a walk through the ocean of most souls Alien | would scarcely get your feet wet. - Deteriorata --------| decvax!frog!cpoint!alien bu-cs!mirror!frog!cpoint!alien
davidbe@sco.COM (The Cat in the Hat) (12/01/89)
news.admin's own jeffd@ficc.uu.net (jeff daiell) said:
-
-All of the proposals for new voting schemes -- and they seem
-to be getting hairier by the moment -- ignore one fact:
-the status quo works pretty well. An occasional abuse,
-like con.aquaria, does happen, but I'd hate to see the
-current nature of the net abandoned in an overly-zealous
-reach to stamp out all possible problems.
Hear hear!
STV or Mauve or whatever may be nice, but it doesn't change the fact that
it's fixing something that's mostly not broken. Sure, the current system
needs a little adjustment (what's the status on the 2/3 yes vote idea,
anyway?) but that doesn't mean it's broken.
The problem with Mauve is that enough people can vote without any of them
getting enough of a majority to show "the will of the net". This has been
shown a number of times.
The problem with STV is complexity. I've yet to see anyone give a clear
explanation of how to tally such a vote. All the explanations people have
given so far look like something the defense department came up with.
And the problem common to *both* proposals, is that they give "no" and
no votes much less significance.
Remember: any voting scheme should be understandable by a 3 year-old.
Because even without talk.kids, there are still people on the net who
won't understand anything more (not naming names, nosiree bob, uh uh...).
--
David Bedno, Systems Administrator, The Santa Cruz Operation, Inc.
Email: davidbe@sco.COM / ..!{uunet,sun,ucbvax!ucscc,gorn}!sco!davidbe
Phone: 408-425-7222 x5123 Disclaimer: Speaking from SCO but not for SCO.
"Can you keep a secret?"
"I'm a GENTLEMAN. Of *course* I can't keep a secret."
- from Cerebus #125peter@ficc.uu.net (Peter da Silva) (12/01/89)
In article <479@scorn.sco.COM> davidbe@sco.COM (The Cat in the Hat) writes: > The problem with STV is complexity. I've yet to see anyone give a clear > explanation of how to tally such a vote. All the explanations people have > given so far look like something the defense department came up with. I'll describe a procedure. It's longer than 25 lines, but it does contain two shell scripts. First of all, set up a file that looks like so: work-file: ------ name name name * user ... ------ Now run the following script: step: ------ /bin/awk '{ votes[$1]++ } END { for(i in votes) print votes[i], i }' work-file | sort -n ------ You now have a list of the names in order, with the fewest votes first. If the last name in this list wins, you're done. Otherwise, take the first name on the list other than '*' or 'NO' and feed it to this script: eliminate: ------ ed - work-file << EOF g/^$1 /s/// w q EOF ------ Now go back and run 'step'. If you have a winner (>50% of the votes, more than 100 votes over the NO vote, etc...), you're done. Otherwise go back and run 'eliminate' again. Keep it up until you've got a winner or the vote has failed. > And the problem common to *both* proposals, is that they give "no" and > no votes much less significance. Funny, there are other people complaining that STV gives "no" *more* significance because it splits the "yes" vote. You can't win. > Remember: any voting scheme should be understandable by a 3 year-old. "Pick the name you like best. OK, put it here. Now, Jimmy, let's go back to the pile. What do you like best..." The voting part. I wouldn't expect a 3-year-old to *run* a vote. -- `-_-' Peter da Silva <peter@ficc.uu.net> <peter@sugar.lonestar.org>. 'U` -------------- +1 713 274 5180. "The basic notion underlying USENET is the flame." -- Chuq Von Rospach, chuq@Apple.COM
gil@banyan.UUCP (Gil Pilz@Eng@Banyan) (12/01/89)
In article <11832@cbnews.ATT.COM> wbt@cbnews.ATT.COM (William B. Thacker,00440,cb,1D211,6148604019) writes: >In article <7139@ficc.uu.net> jeffd@ficc.uu.net (jeff daiell) writes: >>All of the proposals for new voting schemes -- and they seem >>to be getting hairier by the moment -- ignore one fact: >>the status quo works pretty well. An occasional abuse, >>like con.aquaria, does happen, . . . . >Hear ! Hear ! [stuff removed] >Let's consider the actual cost of the sci.aquaria debate: [gory details removed] >Was it worth it ? What was gained ? . . well it was a lot of fun anyway . . I don't get your point Bill, should we change the "voting" guidelines or not ? First you agree with Jeff that the status quo works well enough and then you tell us how badly it fucked up. All these new voting mechanisms are supposed to _fix_ the above problem(s). Why not give 'em a chance ? Do you think we won't be able to change back to the old mechanisms if the new ones don't work out ? p.s. I am very, very sorry for ever voting against sci.aquaria. I had a long talk with Eris and she was extremely pissed at me. I don't know what I could have been thinking of, arguing for "order" in/on the net. I will now vote YES for anything. Gilbert W. Pilz Jr. gil@banyan.com
brad@looking.on.ca (Brad Templeton) (12/02/89)
Compared with other voting-like systems, the current voting like system is no more or less broken, that's true. But I would like to see the "it isn't broken" advocates come out with some evidence to support the claim that this system *is* working. Yes, something gets done. And yes, on the odd votes that fail, the champion sometimes (but not always) goes away quietly. But is it picking the right groups? We are looking for a system to answer the question, "is this group valuable enough that it should, by default, be distributed everywhere?" That's the question being asked. I am more and more convinced the current system measures only the enthusiasm of the group's champion. And not just in cases like sci.age.of.aquarius. What a dumb thing to measure. My examination of readership from sites that report it says that this system has done rather poorly. It has created very few groups that have reached the highest levels of readership. And I would even venture that the groups it has created that did reach those levels would have been created by just about any system. The "alt" random system has done just about as well. So come on, folks. Instead of saying "me, too," show us some *evidence* for your claim that this voting farce works. -- Brad Templeton, ClariNet Communications Corp. -- Waterloo, Ontario 519/884-7473
emv@math.lsa.umich.edu (Edward Vielmetti) (12/02/89)
Brad, the voting farce works. For the conceptually simple groups with easy names to pick and reasonable expectation of interest, it works quite nicely. e.g. soc.culture.korean For groups which have a slightly more difficult to pick out name it seems to work OK, as long as the name-picker person is a calm, non-contoversial, patient person. e.g. rec.radio.shortwave For groups with a lot of interest that appears in a short time, it's horrible -- but that's what alt is for. e.g. alt.sys.next, alt.fusion. For groups which I've been waiting since boo to be created, and which really should just be created (damn the vote, it's about time, what are you waiting for) it creates some unnecessary delays -- but that's what alt is for. e.g. comp.lang.perl (I was about to type comp.sys.perl there :-). For groups which don't fit neatly into the hierarchy as is, it's rotten. I submit any method of deciding the names of these groups is going to be difficult -- but that's what alt is for. e.g. "soc.sex", sci.aquaria. For groups with champions who are net.personaliites, it's rotten. But that's what (alt, biz, clari) are for. e.g. alt.aquaria, biz.clari.sample. For groups which are reasonably happy as alt.xxx and want to switch over to comp,sci,rec,soc,talk,news,misc, it would seem to be of mixed usefulness. sci.physics.fusion took a long time, there's no easy name for alt.fax, comp.sys.next was newgrouped without a vote, and rec.pets.fish never made it. I think that the voting scheme has been a good experiment. Now I'd like to see a change to it. First is to force the group champion to come up with a one line group description as part of the call for discussion, without tying this to a name. Let the discussion ensuing discover the right name, and only then let the call for votes go. --Ed
bill@twwells.com (T. William Wells) (12/03/89)
In article <55499@looking.on.ca> brad@looking.on.ca (Brad Templeton) writes:
: My examination of readership from sites that report it says that this system
: has done rather poorly. It has created very few groups that have reached the
: highest levels of readership. And I would even venture that the groups it
: has created that did reach those levels would have been created by just about
: any system. The "alt" random system has done just about as well.
Brad, stop this nonsense; remember:
readership level != newsgroup quality.
Before we can even debate whether the current system "works", we
have to define what "works" means. And it certainly does *not*
mean "creating newsgroups with wide readership".
The definition of "works" has the same problem that all attempts
at defining a collective "good" have. It can't be done, except
with reference to individual "good"s. Similarly, the system
"works" only and to the extent that those who use the system feel
that it works.
Given the continually growing nature of the net, I'd say that it
works. Given that most "votes" are accepted, I'd say that the
current "voting" system "works".
---
Bill { uunet | novavax | ankh | sunvice } !twwells!bill
bill@twwells.combrad@looking.on.ca (Brad Templeton) (12/03/89)
In article <1989Dec2.213242.12967@twwells.com> bill@twwells.com (T. William Wells) writes: >Brad, stop this nonsense; remember: > > readership level != newsgroup quality. > >Before we can even debate whether the current system "works", we >have to define what "works" means. And it certainly does *not* >mean "creating newsgroups with wide readership". No, readership level does not necessarily equal newsgroup quality. But the most common theme I hear people espouse here is that the "voting" system is there to measure interest in a newsgroup. Well readership level does == interest in the newsgroup. If we could get an accurate measurement of readership level (and improving the accuracy of the current efforts is something that's underway) then we do measure what most people have said they want to measure. We aren't out to measure "quality" of groups here, anyway. Quality is some highly subjective thing that can't easily be quantified. Nor is it the right thing to measure. There are some groups that I am sure have high quality that I do not read or even feed to this site. In order to decide the question "Should this group, be default, be distributed over the whole net?" is to measure the utility of the group, not the quality. I read the groups that have the highest utility -- I get more out of them (for whatever reasons -- "quality", technical info, etc.) than they cost to feed and read. What other standard can there be? Can anybody seriously suggest that large numbers of groups be fed to sites where nobody wants to read them? So readership level is == group quality, if you take quality to mean value to the readers. And it is the readers that count. Each message is posted once, but scanned, skipped or read by up to 20,000 readers. The poster just doesn't even enter into the equation. -- Brad Templeton, ClariNet Communications Corp. -- Waterloo, Ontario 519/884-7473
russotto@eng.umd.edu (Matthew T. Russotto) (12/04/89)
In article <55499@looking.on.ca> brad@looking.on.ca (Brad Templeton) writes: >Compared with other voting-like systems, the current voting like system >is no more or less broken, that's true. > >My examination of readership from sites that report it says that this system >has done rather poorly. It has created very few groups that have reached the >highest levels of readership. So? How do you define 'highest levels of readership'? Using a percentile system? > >So come on, folks. Instead of saying "me, too," show us some *evidence* for >your claim that this voting farce works. What's your evidence that it doesn't? It seems to me that basically groups that are worthwhile to large minorities of people on the net are being created, controversial or not-- Yes, few of these have entirely general interest (how many DEC users read comp.sys.ti?), but is entirely general interest possible with a net this large. >-- >Brad Templeton, ClariNet Communications Corp. -- Waterloo, Ontario 519/884-7473 And BTW, what is your alternative? -- Matthew T. Russotto russotto@eng.umd.edu russotto@wam.umd.edu ][, ][+, ///, ///+, //e, //c, IIGS, //c+ --- Any questions?
wbt@cbnews.ATT.COM (William B. Thacker) (12/04/89)
In article <623@banyan.UUCP> gil@banyan.com writes: >In article <11832@cbnews.ATT.COM> wbt@cbnews.ATT.COM (William B. Thacker,00440,cb,1D211,6148604019) writes: >>Let's consider the actual cost of the sci.aquaria debate: > >[gory details removed] > >>Was it worth it ? What was gained ? > >I don't get your point Bill, should we change the "voting" guidelines >or not ? First you agree with Jeff that the status quo works well >enough and then you tell us how badly it fucked up. I probably didn't make myself clear. My point was that, while what sparked the uproar was Richard Sexton calling the vote for sci.aquaria, whereas many people thought it should have been rec.aquaria, the uproar itself only happened because *people refused to abide by the guidelines then in existance*." Rather than let the vote run its course, everyone insisted on trying to abort it immediately. Various complaints I heard were: - Vote called for before the discussion period had ended - Vote called for without resolution of name issue - Sexton actively solicited votes from "uninterested" parties - etc During the vote, these and other issues turned news.groups into the most boring flamefest I've seen yet. Worse, even, than talk.politics.misc. Well, close, anyway. Kaldis never posted to news.groups. Anyway, all of these flames had nothing to do with, and would not be solved by, the fancy new voting proposals now being considered. None of these will solve the basic factor that allowed this whole affair ; that Usenet is an anarchy, and if somebody wants to ignore the rules, they can. Richard chose to violate certain guideline procedures (*), and the *violation* cased 6 weeks of screaming. Writing new procedures won't prevent him, or anyone else, from violating those, too, with the same results. (*) Some will argue whether guidelines were, in fact, violated. I won't. If you prefer, consider that the *perceived* violation of guidelines caused the arguments. Same effect... You could institute a rule saying "New groups will only be created if they are approved by Greg Woods, and no voting will be allowed or adhered to," and it won't prevent the next person from coming along and conducting a vote, anyway, while 50 people yell about the violation of the guidelines. We can't enforce the guidelines we have, so what's the point of writing new ones ? Now, I'm not claiming that it's therefore impossible to control new group creation. Rather, what I'm driving at is this: Stick to the guidelines we have, and leave the ultimate choice for group creation where it *always* was; in the hands of the administrator of each Usenet site. Remember, the votes are simply suppose to show support for the group, and opposition to it. They're not binding to the site admins; rather, they demonstrate to the admins that the group is worth creating. I think Richard's vote still contained that information; the huge "yes" vote indicated (*to me*) a bit of shady doings, but nonetheless a sincere interest in the new group, while the huge "no" vote indicated (*to me*) equally questionable turnout, and a sincere objection to the choice of name. That's all the information a site administrator needs to decide. Why does everyone insist on coming up with these new contrivances to make the admin's choice for him ? Don't you think she's capable of deciding for herself ? >All these new >voting mechanisms are supposed to _fix_ the above problem(s). Why not >give 'em a chance ? Because they won't fix the problem. The problem is not the creation of sci.aquaria; it was the weeks of anguish it took to get it over with. I believe that, had everyone just quietly cast their votes, rather than screaming back and forth in news.groups, it might have failed; it certainly would not have passed any more convincingly. So the result would be about the same; a narrow margin with several hundred no votes. And the site admins would have looked at it and said, "well, it made the 100 vote margin, but I think the name's stupid, and lots of people agree, so *I* won't carry it." Or decided to carry it, despite the large negative vote. Which is what we have now, except we did it the hard way. >Do you think we won't be able to change back to >the old mechanisms if the new ones don't work out ? No, I don't think we'll ever go back. That's not the way a bureaucracy works. - - - - - - - - valuable coupon - - - - - - - clip and save - - - - - - - - Bill Thacker AT&T Network Systems - Columbus wbt@cbnews.att.com "C" combines the power of assembly language with the flexibility of assembly language.
wbt@cbnews.ATT.COM (William B. Thacker) (12/04/89)
In article <55499@looking.on.ca> brad@looking.on.ca (Brad Templeton) writes: > >But I would like to see the "it isn't broken" advocates come out with >some evidence to support the claim that this system *is* working. Usenet continues to grow. People aren't giving it up in droves. *Somebody* sure seems to think it's doing well. >Yes, something gets done. And yes, on the odd votes that fail, the champion >sometimes (but not always) goes away quietly. My contention is that, whatever the guidelines, the "rude champion" will continue to be a problem. >But is it picking the right groups? We are looking for a system to >answer the question, "is this group valuable enough that it should, by >default, be distributed everywhere?" That's the question being asked. Do you still stick with your earlier claim that a group should average at least one reader per site according to arbitron ? I have never agreed with this belief. >My examination of readership from sites that report it says that this system >has done rather poorly. It has created very few groups that have reached the >highest levels of readership A useless statement. It is a mathematical axiom that no more than 5% of the sample can be in the top 5%. Now, if you're saying that the range of interest (from 60,000+ for the number 1 group to 3000 or so for the smallest group) is too great, I see your point; but I don't agree with it. I guess I don't understand. Are you saying that there are many potentially widely-read groups that are somehow being suppressed by the current guidelines ? If so, why haven't any of the tens of thousands of interested parties proposed them ? - - - - - - - - valuable coupon - - - - - - - clip and save - - - - - - - - Bill Thacker AT&T Network Systems - Columbus wbt@cbnews.att.com "Kaldis seems to be one of the few rational people on this net ... " - Jack Schmidling
dg@lakart.UUCP (David Goodenough) (12/05/89)
brad@looking.on.ca (Brad Templeton) sez: > But is it picking the right groups? We are looking for a system to > answer the question, "is this group valuable enough that it should, by > default, be distributed everywhere?" That's the question being asked. > ..... > My examination of readership from sites that report it says that this system > has done rather poorly. It has created very few groups that have reached the > highest levels of readership. One possible cause of the above is Harris' lament: "all the good ones are taken". Just suppose that comp.lang.c didn't exist, and someone proposed it. Under the current system it would probably be a landslide, and get propagated everywhere. But why will we never see this? Because it already exists. Partly by nature of the net, all the _REALLY_ popular newsgroups have already been created, and the ones coming along now are not going to get as wide coverage. Very probably soc.culture.korean will not get the readership that comp.unix.wizards does. Just because it plays to a smaller audience, is it any less a newsgroup? -- dg@lakart.UUCP - David Goodenough +---+ IHS | +-+-+ ....... !harvard!xait!lakart!dg +-+-+ | AKA: dg%lakart.uucp@xait.xerox.com +---+
alien@cpoint.UUCP (Alien Wells) (12/05/89)
In article <11941@cbnews.ATT.COM> wbt@cbnews.ATT.COM (William B. Thacker,00440,cb,1D211,6148604019) writes: >In article <55499@looking.on.ca> brad@looking.on.ca (Brad Templeton) writes: >> >>But I would like to see the "it isn't broken" advocates come out with >>some evidence to support the claim that this system *is* working. > >Usenet continues to grow. People aren't giving it up in droves. >*Somebody* sure seems to think it's doing well. Just because Usenet is a valuable and respected service does not mean that it doesn't have problems or that some facets of it aren't working. An earlier posting claimed the problem in the sci.aquaria debate wasn't the choice of the 'wrong' name (I disagree ...) but the flaming disagreement that the debate brought out (I have to agree with the sentiment, though). However, I stongly disagree that 'fixing' things wouldn't have tempered the flames. Consider the following: - A lot of the flaming was brought out by people sincerely disagreeing about what the 'right' name ought to be, and Usenet voting gives them no other method of objection but to flame. If name A wins by a 2x margin, it really doesn't matter whether name B would have won by a 10x margin. If you can't get enough votes to kill a bad name, you have to either live with it or start pulling 'my site isn't going to honor your vote' tantrums. A multiple voting scheme, MAUVE, STV, or WEIP, would give people a way to say which name they want. - A lot of the 'dishonorable campaigning' was caused by people (in both camps) going out and saying 'Hey, send this vote to Sexton (and/or else) this will happen'. If a multiple vote was being held, it would be harder to keep the whole story away from potential voters. For instance, someone who just wants an aquarium group would see that he is voting NO to a number of names and wonder why. The sci.aquaria mess shows clearly that the process IS broken. We have a case where a name passed even though it is clear (to almost everyone) that rec.aquaria would have had a lot more support and less opposition. There were vastly more flames than I would have liked to have waded through. And now we have a case where the result of the vote is being ignored by a lot of system admins and the group is a joke. Multiple voting would have prevented the group champion from forcing an unpopular name on the net. Multiple voting would have negated the perception of a power play. Multiple voting would have given people a 'non-violent' way to express their disapproval. Multiple voting would have 'fixed' this problem. And before you suggest this is an isolated incident, look at the .pagan and sci.groupware debates. I hope they get resolved better than sci.aquaria, but there isn't any support from net guidelines that would stop them from turning into another fiasco. Look at sci.skeptic and comp.women. Similar things happened. And consider what would happen if another 'great renaming' were to be contemplated. Without multiple voting, it would be a mess where the net would be (to a large extent) at the mercy of the renaming champions. The system IS broken. Take a look at some of the rather violent reactions I got when I tried to fix the .aquaria problem ... -- --------| Rest assured that a walk through the ocean of most souls Alien | would scarcely get your feet wet. - Deteriorata --------| decvax!frog!cpoint!alien bu-cs!mirror!frog!cpoint!alien
brad@looking.on.ca (Brad Templeton) (12/05/89)
In article <11941@cbnews.ATT.COM> wbt@cbnews.ATT.COM (William B. Thacker,00440,cb,1D211,6148604019) writes: >In article <55499@looking.on.ca> brad@looking.on.ca (Brad Templeton) writes: >> >>But I would like to see the "it isn't broken" advocates come out with >>some evidence to support the claim that this system *is* working. > >Usenet continues to grow. People aren't giving it up in droves. >*Somebody* sure seems to think it's doing well. Do you suggest USENET grows because we have such a nice voting system to pick groups with? Sorry, but I fail to see how you drew this conclusion. Anecdotal evidence does say people have left USENET in droves. Of the active netters of the early days, only a very few still take an active role. At places like Usenix and Unix expo I regularly get the chance to ask people if they read USENET, and I very frequently get the answers, "used to," "too noisy" or "don't have the time." Brian's stat, "Percentage of users who are netreaders" isn't all that meanigful, but it has dropped about 2-3% this year. > >Do you still stick with your earlier claim that a group should average at >least one reader per site according to arbitron ? I have never agreed >with this belief. Actually, I would probably require 1/2 or 1/3 reader per site, but I don't know how you can disagree with this, given the assumption of accurate readership measurement. The alternative is to suggest we send groups to thousands of machines where nobody reads them. The question that the guidelines try to answer, don't forget, is, "what groups by default should go to all machines." NOT, "which are the best groups." NOT, "which are the most interesting." NOT, "which have the most technical value." At the very least we should not run around picking groups that will not be read on most machines! > >>My examination of readership from sites that report it says that this system >>has done rather poorly. It has created very few groups that have reached the >>highest levels of readership > >A useless statement. It is a mathematical axiom that no more than 5% of >the sample can be in the top 5%. My analysis said it was far worse than that, and in particular, if you recall, that "alt" groups and groups created against the guidelines (a "control base" if you will) had as good a track record of getting good readership. >I guess I don't understand. Are you saying that there are many >potentially widely-read groups that are somehow being suppressed by the >current guidelines ? If so, why haven't any of the tens of thousands of >interested parties proposed them ? Yes. For example many groups that get created in alt. Many groups that don't get created because nobody wants the messy, time-consuming job of running a net vote. Remember Brian Reid's post of a while ago. "Here's a group that would be of interest, but I have no desire to go through the silly guideline procedure. Would somebody else please do that for me?" He was lucky, Salz did. But what about splitting ibm.pc? It's been suggested at least a dozen times, been done to a degree, in alt, but nobody has seen it through. -- Brad Templeton, ClariNet Communications Corp. -- Waterloo, Ontario 519/884-7473
brad@looking.on.ca (Brad Templeton) (12/05/89)
In article <11939@cbnews.ATT.COM> wbt@cbnews.ATT.COM (William B. Thacker,00440,cb,1D211,6148604019) writes: >Anyway, all of these flames had nothing to do with, and would not be solved >by, the fancy new voting proposals now being considered. >.... >We can't enforce the guidelines we have, so what's the point of writing new >ones ? Quite right -- there is no point in adding new rules, or more complex rules. This does not, however, invalidate the concept of a whole new set of rules. The trial newsgroup concept, as proposed by myself and others, goes in a completely different direction. And while you may not like some parts of it, it does not have the problem of discussion. There is no discussion at all. If we used trial newsgroups we could rmgroup news.groups and never miss it. No argument because there is nothing to argue about. That's the only way an anarchy can really work. You can never have a working anarchy by "voting." Remember the goals a newsgroup creation scheme should work towards: a) It must provide some sort of objective demonstration of how widely desired the group is. b) It must shut up the group champion if the group is not to be created. b.1) It helps if it has the illusion of democracy. c) It should be efficient, wasting a minimum of net admins time on pointless discussion, and it should discourage discussion of things that simply aren't worth discussing outside the new group itself. d) It should be fast where possible. e) It should pick consistent names. Trial groups perform excellently except on "e" which is an independent problem. I don't think the various suggestions that involve voting on names are good solutions at all. The current system does well only at b and b.1. But it is failing those, too. -- Brad Templeton, ClariNet Communications Corp. -- Waterloo, Ontario 519/884-7473
wbt@cbnews.ATT.COM (William B. Thacker) (12/05/89)
In article <56804@looking.on.ca> brad@looking.on.ca (Brad Templeton) writes: >In article <11941@cbnews.ATT.COM> wbt@cbnews.ATT.COM (William B. Thacker,00440,cb,1D211,6148604019) writes: >>In article <55499@looking.on.ca> brad@looking.on.ca (Brad Templeton) writes: >>> >>>But I would like to see the "it isn't broken" advocates come out with >>>some evidence to support the claim that this system *is* working. >> >>Usenet continues to grow. People aren't giving it up in droves. >>*Somebody* sure seems to think it's doing well. > >Do you suggest USENET grows because we have such a nice voting system >to pick groups with? No, I merely indicate that whatever's wrong with the voting system can't be *that* bad, because it still results in a popular product. Of course, this doesn't mean it can't be better; simply that it's not all that bad. And I'm not inclined to change it until something comes along that will clearly solve the few problems the current system has. None of the proposals I've seen yet would do so. > Sorry, but I fail to see how you drew this conclusion. >Anecdotal evidence does say people have left USENET in droves. Apparently, then, for every one who leaves, two more step in to take her place. >Of the >active netters of the early days, only a very few still take an active >role. At places like Usenix and Unix expo I regularly get the chance to >ask people if they read USENET, and I very frequently get the answers, >"used to," "too noisy" or "don't have the time." So the fact that I no longer play my euphonium in a band indicates a fundamental flaw in instrumental music ? (Or perhaps just *my* music 8-) Interests come and go. >>Do you still stick with your earlier claim that a group should average at >>least one reader per site according to arbitron ? I have never agreed >>with this belief. > >Actually, I would probably require 1/2 or 1/3 reader per site, but I >don't know how you can disagree with this, given the assumption of >accurate readership measurement. The alternative is to suggest we >send groups to thousands of machines where nobody reads them. Why not ? I've been through this before, but I'll try again. Let's say I'm the sysadmin on fubar.hal.com, and I, or one of my readers, want to read rec.food.artichokes, a small group with only a few thousand estimated readers. Meanwhile, you're the sysadmin of tarfu.buu.com, and have an interest in talk.glue, another small group. Neither of us in intersted in the other's group. I want r.f.a to get wide propogation, so my reader will get the most use of it. That means, among other things, that I want *you* to propogate it to all the sites that feed from you, as they may also have interested readers. Meanwhile, you would like the same treatment for t.g. I therefore agree to carry and forward t.g if you'll do the same for r.f.a. In return for carrying a group you don't read, you get an improvement in the group you *do* read. Likewise for me. Cooperation for mutual benefit. Where this begins to break down is when, after a few months, you realize that the volume in r.f.a is 20 times that in t.g. You're incurring much higher expenses on my account than I am on yours. So it seems to me that volume per reader is a much better indicator of a group's "worthiness" than is number of readers. >The question >that the guidelines try to answer, don't forget, is, "what groups by >default should go to all machines." It seems to me that, rather, the guidelines are to demonstrate that there is sufficient interest in the group to verify that sites you feed are likely to have interested readers, and enough of them to warrant your carrying of the group. They don't equate. A "backbone" site might feed several thousand sites (a few dozen directly, each of which feed a few dozen, etc). Obviously, these sites are much more likley to find interested readers among their "dependents" than would a site that only feeds two or three machines. And in the extreme, a leaf site obviously has no need to carry anything not read by its readers. Therefore, there's no system of voting that can really say "what groups should by default go to all machines;" that choice is left to the sysadmin. >>>My examination of readership from sites that report it says that this system >>>has done rather poorly. It has created very few groups that have reached the >>>highest levels of readership >> >>A useless statement. It is a mathematical axiom that no more than 5% of >>the sample can be in the top 5%. > >My analysis said it was far worse than that, and in particular, if you >recall, that "alt" groups and groups created against the guidelines >(a "control base" if you will) had as good a track record of getting >good readership. Which tells me that the current guidelines are too restrictive. It does not seem to indicate that we need yet further restrictions. - - - - - - - - valuable coupon - - - - - - - clip and save - - - - - - - - Bill Thacker AT&T Network Systems - Columbus wbt@cbnews.att.com Free the Lagrange 5 !
brad@looking.on.ca (Brad Templeton) (12/07/89)
In article <11984@cbnews.ATT.COM> wbt@cbnews.ATT.COM (William B. Thacker,00440,cb,1D211,6148604019) writes: >>The question >>that the guidelines try to answer, don't forget, is, "what groups by >>default should go to all machines." > >It seems to me that, rather, the guidelines are to demonstrate that >there is sufficient interest in the group to verify that sites you >feed are likely to have interested readers, and enough of them to warrant >your carrying of the group. So what you're saying is that rather than measure how many interested readers there *actually are*, we should try to measure if the group is "likely to have interested readers," and that we should do this by checking if 150 people (typical number) will send email supporting the group? Seems to me that creating the group in a limited area and checking the real interest there is 100 times better than trying to find out if interest is likely through a complex and noisy procedure. To answer another question -- a good metric would take some account of net topology. For example, if you have a reader downstream that you are willing to feed, that counts as having a reader on your site. To calculate this, all we have to do is flow the readership data up backwards through the flow graph, or rather the directed subset of that graph. -- Brad Templeton, ClariNet Communications Corp. -- Waterloo, Ontario 519/884-7473
learn@igloo.scum.com (Bill Haroldmegrhondapooh Vajk) (12/07/89)
In article <11984@cbnews.ATT.COM> William B. Thacker writes: > No, I merely indicate that whatever's wrong with the voting system can't be > *that* bad, because it still results in a popular product. Things couldn't have been bad all those years in East Germany, cause people stayed. Give them a little freedom, and they run away from it. Kinda proves freedom is a bad thing for East Germans, dontcha think ? Bill.etc | It is not in the nature of reason to regard | things as contingent, but as necessary. - Spinoza
brad@looking.on.ca (Brad Templeton) (12/07/89)
In article <765@lakart.UUCP> dg@lakart.UUCP (David Goodenough) writes: >Very probably soc.culture.korean will not get the readership that >comp.unix.wizards does. Just because it plays to a smaller audience, is it >any less a newsgroup? No, not when comparing a group with 1 reader/machine to a group with 6 readers/machine. But there has to be a line somewhere, where a group is less worthy of complete-net-propagation if very few want to read it. To paraphrase a joke from RHF, soc.culture.jewish.lesbian.civil_war_daughters might have an audience, but it shouldn't be a newsgroup. I am all in favour of drawing the line as low as we can as resources get cheaper and cheaper. But it does have to be drawn somewhere. Topics below that line simply have to become mailing lists, or groups with dynamic distribution. -- Brad Templeton, ClariNet Communications Corp. -- Waterloo, Ontario 519/884-7473
rissa@attctc.Dallas.TX.US (Patricia O Tuama) (12/07/89)
In article <56882@looking.on.ca> brad@looking.on.ca (Brad Templeton) writes: >There is no discussion at all. If we used trial newsgroups we could >rmgroup news.groups and never miss it. No argument because there is >nothing to argue about. That's the only way an anarchy can really work. >You can never have a working anarchy by "voting." Gee, I don't know, Brad, -- sci.aquaria was on trial for three years and still there was a minor disagreement regarding it's creation as I recall. at that point trish, remembering the good old days of nets and mods .
wbt@cbnews.ATT.COM (William B. Thacker) (12/08/89)
In article <57577@looking.on.ca> brad@looking.on.ca (Brad Templeton) writes: >In article <11984@cbnews.ATT.COM> wbt@cbnews.ATT.COM (William B. Thacker,00440,cb,1D211,6148604019) writes: >>>The question >>>that the guidelines try to answer, don't forget, is, "what groups by >>>default should go to all machines." >> >>It seems to me that, rather, the guidelines are to demonstrate that >>there is sufficient interest in the group to verify that sites you >>feed are likely to have interested readers, and enough of them to warrant >>your carrying of the group. > >So what you're saying is that rather than measure how many interested >readers there *actually are*, we should try to measure if the group is >"likely to have interested readers," and that we should do this by >checking if 150 people (typical number) will send email supporting the >group? No. I'm saying that's what we do now, and have always done since these guidelines have been in effect. And, since I feel the current system is OK, my opinion is that the guidelines should continue to have this meaning. As I said, this is based on my opinion that the newsgroups currently being created are, by and large, okey-dokey. >Seems to me that creating the group in a limited area and checking >the real interest there is 100 times better than trying to find out if >interest is likely through a complex and noisy procedure. Given a spherical football, yes. What if the limited area is not really representative of the entire net ? For example, create a trial group for discussing Hispanic culture in each of the following spots: New York, Texas, Germany, Nebraska, Wisconsin, and Florida. Other than that, I've not followed your Trial Newsgroup proposal, so I'd best not comment. - - - - - - - - valuable coupon - - - - - - - clip and save - - - - - - - - Bill Thacker AT&T Network Systems - Columbus wbt@cbnews.att.com Black Holes are where God is dividing by zero
peter@ficc.uu.net (Peter da Silva) (12/08/89)
From my .signature: "You can tell when a USENET discussion is getting old when one of the participants drags out Hitler and the Nazis" -- Richard Sexton Looks like there's a new variant: In article <1475@igloo.scum.com> learn@igloo.scum.com (Bill Haroldmegrhondapooh Vajk) writes: > Things couldn't have been bad all those years in East Germany, cause > people stayed. Give them a little freedom, and they run away from it. I think this discussion has gone on too long. -- `-_-' Peter da Silva. +1 713 274 5180. <peter@ficc.uu.net>. 'U` Also <peter@ficc.lonestar.org> or <peter@sugar.lonestar.org>. "If you want PL/I, you know where to find it." -- Dennis
sloane@kuhub.cc.ukans.edu (12/09/89)
In article <57749@looking.on.ca>, brad@looking.on.ca (Brad Templeton) writes: > I am all in favour of drawing the line as low as we can as resources get > cheaper and cheaper. But it does have to be drawn somewhere. Topics > below that line simply have to become mailing lists, or groups with > dynamic distribution. So what is the cutoff point? I have heard you mention 1 reader for every 2 machines. The latest version of the arbitron survey shows an estimated 18981 sites on usenet. Does this mean that any group with less than about 9000 readers estimated should be converted to a mailing list? I suspect that a mailing list with 9000 subscribers would cost a lot more than a newsgroup, even if you don't count the administrative time it would take up. Does anyone know what the break even point is, that is, when does a mailing list become more economical than a newsgroup? -- USmail: Bob Sloane, University of Kansas Computer Center, Lawrence, KS, 66045 E-mail: sloane@kuhub.cc.ukans.edu, sloane@ukanvax.bitnet, AT&T: (913)864-0444
welty@lewis.crd.ge.com (richard welty) (12/09/89)
In article <19468.257f9076@kuhub.cc.ukans.edu>, sloane@kuhub.cc.ukan writes:
*Does anyone know what the break even point is, that is, when does a mailing
*list become more economical than a newsgroup?
it seems to me that about a year ago, Chuq made an educated guess
that it was in the vicinity of 250 or 300 readers. personally,
i find that handling mailing lists of over 100 people are a real
administrative headache (all the lists i currently run are much,
much smaller than this.) some are less bothered by such a load
than i, though.
richard
--
richard welty 518-387-6346, GE R&D, K1-5C39, Niskayuna, New York
..!crdgw1!lewis.crd.ge.com!welty welty@lewis.crd.ge.com
``i've got a girlfriend with bows in her hair,
and nothing is better than that'' -- David Byrnebrad@looking.on.ca (Brad Templeton) (12/09/89)
In article <10512@attctc.Dallas.TX.US> rissa@attctc.Dallas.TX.US (Patricia O Tuama) writes: >In article <56882@looking.on.ca> brad@looking.on.ca (Brad Templeton) writes: >>There is no discussion at all. If we used trial newsgroups we could >>rmgroup news.groups and never miss it. No argument because there is >>nothing to argue about. That's the only way an anarchy can really work. >>You can never have a working anarchy by "voting." > >Gee, I don't know, Brad, -- sci.aquaria was on trial for three years >and still there was a minor disagreement regarding it's creation as >I recall. Sci.aquaria was not on trial, it was in alt. It was not doing super well in alt, at just over .56 readers/site, but that exceeds my proposal for a creation requirement of .5/site. So, under my proposal, alt.aquaria could have become a netwide group without any discussion. The choice of name is a somewhat orthoganal issue. Ideally name and distribution would be independent and this would not have been a problem. But that aside, current schemes don't handle the name question and would not have done any better. My proposal is that a group of 1-3 people pick the name as a service. That would have solved the problem as long as the people knew enough to ignore whiners. You ignore the whiners on the net and they go away soon enough. -- Brad Templeton, ClariNet Communications Corp. -- Waterloo, Ontario 519/884-7473
brad@looking.on.ca (Brad Templeton) (12/09/89)
We come back to the original question. People have various opinions about how well the process has worked. Perhaps it is important, before proposing any method, to define some sort of metric for the success of that method. Admittedly with some schemes the metric *is* the method, as might appear to be the case for readership measurements. But with trial groups, it would be worthwhile comparing readership from the trial period and after creation (for successes) to see how good a predictor the trial distribution was. And yes, the trial distribution could be skewed. But not nearly as much as "voting" is skewed! And those who feel they are not represented are thus encouraged to send in reports. If they do it regularly, then not only do we get their slightly skewing reports for the trial group at hand, but their non-biased reports for the other trial groups, and trial groups to come. I expect most of trial would get at least a 40% propagation -- perhaps more. Alt gets as much as 60 to 70%. Two special sub-hierarchies, namely trial.dangerous (for groups like sex, drugs etc. that might be illegal in some jurisdictions) and trial.commercial (for groups like misc.forsale, comp.newprod, news.announce.conferences, misc.jobs.* etc.) would get lower propagation, but not a lot lower, I think. Pretty good results, I predict. Unfortunately, we can hardly find out if people want trial groups over "voting" by holding a "vote!" Such a result would be as meaningful as, "amazingly 100% of survey respondents said that they liked responding to surveys!" -- Brad Templeton, ClariNet Communications Corp. -- Waterloo, Ontario 519/884-7473
rissa@attctc.Dallas.TX.US (Patricia O Tuama) (12/11/89)
In article <58888@looking.on.ca> brad@looking.on.ca (Brad Templeton) writes: >>Gee, I don't know, Brad, -- sci.aquaria was on trial for three years >>and still there was a minor disagreement regarding it's creation as >>I recall. >Sci.aquaria was not on trial, it was in alt. Oh come on, Brad, it's the same concept. >The choice of name is a somewhat orthoganal issue. Ideally name and >distribution would be independent and this would not have been a problem. You know, I don't remember this name biz being such a hassle until recently when all of a sudden it seems to have exploded into inter- stellar prominence. The way some people carry on about this you'd think every single solitary problem usenet encounters could be solved in a second if we just had this little committee to take care of de- ciding what to call new groups. Mark Horton was right. >as the people knew enough to ignore whiners. You ignore the whiners on >the net and they go away soon enough. That's what I kept telling Richard and sure enough, most of them have. yours for nets & mods, . t r i s h a o t u a m a
tjk@csd4.csd.uwm.edu (Thomas Krueger) (12/11/89)
In article <2838147735@lewis.crd.ge.com> welty@lewis.crd.ge.com (richard welty) writes: >In article <19468.257f9076@kuhub.cc.ukans.edu>, sloane@kuhub.cc.ukan writes: >*Does anyone know what the break even point is, that is, when does a mailing >*list become more economical than a newsgroup? > >it seems to me that about a year ago, Chuq made an educated guess >that it was in the vicinity of 250 or 300 readers. personally, >i find that handling mailing lists of over 100 people are a real >administrative headache (all the lists i currently run are much, >much smaller than this.) some are less bothered by such a load >than i, though. > >richard I agree. The list I run is 200 addresses _I_ deliver to, and who knows how many via 25 redistribution lists. It's one of the reasons I'm trying to get a newsgroup started (HINT HINT). - Tom -- "A Veteran of the Psychic Wars" Thomas Krueger UW-Milwaukee College of Engineering Electronics Shop tjk@csd4.csd.uwm.edu [moderator, info-high-audio] +1 414 229 5172