Re: RFD process on hold / Call For Ideas
Mean Green Dancing Machine
Jonathan Kamens <j...@kamens.brookline.ma.us> wrote:
>
>For the record, I'd volunteer to be on either the active committee or
>the advisory committee. I wouldn't do it forever, but I'd take my
>shift.
I'd do the active committee; I'd rather not do the advisory committee,
but I'd probably do a stine if needed.
--
--- Aahz <*> (Copyright 2005 by aa...@pobox.com)
Hugs and backrubs -- I break Rule 6 http://rule6.info/
Androgynous poly kinky vanilla queer het Pythonista
"It's 106 miles to Chicago. We have a full tank of gas, a half-pack of
cigarettes, it's dark, and we're wearing sunglasses." "Hit it."
Russ Allbery
> Many other issues have been brought up in recent discussion in
> news.groups. They include:
I'll throw a few more out there that really concern me and doesn't seem to
be talked about as much:
* Right now, we get very little effective feedback from the people who
actually use Usenet, and therefore very little data from which to
judge whether a system is working or not. We don't really know what
the common use cases are, we don't know what the common confusions
are, and we don't know what people find easy or hard.
* Similarly, we have very few paths of communication to news
administrators, the vast majority of whom do not read news.groups.
We don't know what's working for them and what isn't working for
them. I've tried to contact several key administrators in the past
and get their input, but this ideally would be an ongoing thing
somehow.
* We're not at all proactive about getting a group created (or removed)
on lots of servers. Right now, we're leaving it to the proponent to
do that. This isn't necessarily a horrible decision, but it's not a
great one either, and it really blurs the lines of why one would use
a managed hierarchy rather than just using alt.* if individual effort
to get the group carried is required anyway. I would think that it
would be possible to set up some standing relationships with news
admins and be able to reuse those contacts effectively to get more
ISPs synced up with the list of active groups (or at least understand
better why they're not).
--
Russ Allbery (r...@stanford.edu) <http://www.eyrie.org/~eagle/>
Todd Michel McComb
Russ Allbery <r...@stanford.edu> wrote:
>I would think that it would be possible to set up some standing
>relationships with news admins and be able to reuse those contacts
>effectively to get more ISPs synced up with the list of active
>groups (or at least understand better why they're not).
I think that's an area where we've shown a little bit of improvement
over the past few years, so things are heading (modestly) in the
right direction at least. There's certainly plenty of room for
improvement remaining, of course.
Todd McComb
mcc...@medieval.org
Adam H. Kerman
>mcc...@medieval.org (Todd Michel McComb) writes:
>>>People just aren't willing to expose their e-mail addresses
>>Yes, it is pretty clear that that is an issue. I'm not sure how
>>we make polling work in light of that.
>This is the main problem I was hoping a web-based system would solve.
Tim, I have every confidence that you could come up with a technical
solution. That is a specific problem that can be solved.
We do not have a consensus on what the main problem is.
I think the main problem is (which thus far no one agrees with) lack of
responsible party to identify the Usenet audience for a proposed
newsgroup.
Supposedly, all Usenet users (at least the ones who speak English) should
watch n.a.n for a proposed newsgroup on a topic of interest. I suppose if
they register at your proposed Web site, they could glance at it instead.
In reality, most Usenet users are complacent in these matters. Your Web
site could increase participation if there was some way to ensure that no
one used multiple identities, but any increase in participation would be
small.
>>Personally, I'm thinking we don't want to do away with polling
>>entirely, but we also don't want to rely on it exclusively. I can
>>imagine a "fast track" system that then falls back to polling if
>>the committee has misgivings.
>I absolutely think that this is a good idea; and it seems that a lot of
>others seem to agree. Jonathan Kamens' idea seemed to cover this...
>What kind of consensus do you want to reach on this to decide to work
>with it? Do you want a general scheme? More of the specifics worked
>out? Actual committee members?
All too often, government reorganizes in lieu of addressing a social need
that would require the expenditure of a larger amount of money. It's
easier to build support for the appearance of doing something than
actually doing it.
I hope that's not what we are doing here. Usenet users aren't willing to
follow through on gaining support for groups they allege there's a need
for. No technical committee will do that either.
Todd Michel McComb
Mean Green Dancing Machine <aa...@pobox.com> wrote:
>I'd do the active committee; I'd rather not do the advisory
>committee, but I'd probably do a stine if needed.
If we go this sort of route, it probably makes sense to have the
active committee handle everything group-mentors does right now.
I'm reluctant to post what I'd want to do in such a scheme, so early
in the process, but as long as I'm having the thoughts, I might as
well do that.
First, for one thing, obviously this whole process here this month
(if you can exactly call it a "process") could mean I'm no longer
participating, and I'm totally fine with that. In that sense, what
I think of people's ideas isn't especially important. We want to
come up with the right plan, and in those terms, it doesn't really
matter what I am willing to do or whatever, because I _am_ willing
to step aside. OK, so that's that.
As far as forming a new committee (or two committees) to oversee
group creation and/or removal, I don't think I want to be on either
committee. The following comments assume that Russ is willing to
stay on, as the keeper of the control key, and also as the person
who maintains the isc.org archive and groups list. If that's true,
I'm OK with the following role for me: I would assist Russ in these
areas, I would continue to update the moderators file/list, and I
would serve in an advisory role to the committee(s). The second
part could include leading some "MVI" procedures, if the active
committee finds that worthwhile. The third part would not include
any sort of authority or appeal against the committee. I would
advise them as they wanted, with no ability to fire them, and I
would quite conceivably make their n.a.n posts for them of whatever
their policies and decisions were (or make sure that a suitable
conduit for n.a.n posting existed). This means committee members
could be voted on or off without anyone needing to know how to
pgp-sign the messages. What I would _not_ like is to somehow be
left holding the bag on group creation decisions, but I would
effectively be working _for_ the committee in that role. That part
seems fine to me, and unless people really just want me out of here,
I sure don't see any reason not to continue doing the moderator
updates. Getting that back up to speed was a bit of a task.
Todd McComb
mcc...@medieval.org
Russ Allbery
> I know the uk* is a tiny part of Usenet but we have occasionally
> published "hundreds" of addresses and as far as I know they have never
> been spammed
I don't know why you guys don't get attacked. Unfortunately, we do.
Voting lists from results have definitely been used for spam, multiple
times.
Todd Michel McComb
Russ Allbery <r...@stanford.edu> wrote:
>Voting lists from results have definitely been used for spam,
>multiple times.
Yes, just to clarify, I don't think it's automated spam harvesting
at all. It's malicious people seeking revenge. The voting public
turns out to be our weak link.
Todd McComb
mcc...@medieval.org
Martin X. Moleski, SJ
<dgn9jn$nd0$1...@agricola.medieval.org>:
> ... One alternative is
>to publish addresses exactly as people use them to post, with
>whatever munging they use. We had already discussed doing something
>like this.
That's the beauty of using the experimental groups.
The real names or addresses of the "voters" doesn't matter.
They post as they please--which is the spirit of Usenet.
No need to verify identities.
If someone wants to go to the trouble of building a Potemkin
village of sock puppets to inflate a newsgroup's stats, let 'em.
Trolls like audiences too much to stay all alone in an empty
newsgroup.
Marty
Russ Allbery
> As far as forming a new committee (or two committees) to oversee group
> creation and/or removal, I don't think I want to be on either committee.
> The following comments assume that Russ is willing to stay on, as the
> keeper of the control key, and also as the person who maintains the
> isc.org archive and groups list.
I'm willing to stay on as the keeper of the control key; it's not very
much work, I know how to do it, and I can make it less work if I ever get
around to writing some more automation.
I'm willing to serve in either/both committees for a transition period,
although I'd prefer to phase out my participation in such work after a
year at most, and I have no objections to not being part of the committees
from the beginning.
Maintaining the isc.org archive and groups list is both kind of fun and
really hardly any work at all, so I can continue doing that. I'm not
particularly fast about making updates that require manual intervention,
though.
Walter Roberson
There was also the incident years ago where someone essentially
libelled everyone who voted in favour of the anti-spam newsgroup.
[Or at least it would have been libel if the poster of the defamatory
material had had any credibility.]
--
Entropy is the logarithm of probability -- Boltzmann
Walter Roberson
Adam H. Kerman <a...@chinet.com> wrote:
>Supposedly, all Usenet users (at least the ones who speak English) should
>watch n.a.n for a proposed newsgroup on a topic of interest. I suppose if
>they register at your proposed Web site, they could glance at it instead.
In practice, I rarely check web sites for updates [other than
Google News ;-) ]. I know some browsers have features which allow
one to bookmark sites and automatically check for updates to the
bookmarked sites, but I find I do not use that, and I don't think I'm
likely to start.
n.a.n on the other hand is in my subscribed list, checked several
times per day.
>In reality, most Usenet users are complacent in these matters. Your Web
>site could increase participation if there was some way to ensure that no
>one used multiple identities, but any increase in participation would be
>small.
There's a small problem with that: I use multiple identities, for
non-overlapping purposes. Most things I use my professional identity
for, but my professional life doesn't need the trouble that could
be garnered by my postings to (e.g.) rec.pets.cats.bonzai .
I don't go around sock-puppeting myself, but when it comes time
for voting for a horseradish fetish newsgroup, I'd rather not be
forced to use my single well-known identity.
--
Daylight is a trademark of OSRAM SYLVANIA INC.
edward ohare
Team) wrote:
>However, subsequent discussion raises a more basic question: Does
>the process work at all? It is one thing to change +100 to +50,
>but is this a pointless cosmetic change to a system which has
>outlived its usefulness?
The real problem with the current voting system is it pits two issues
against each other that are not part of the same yes/no question.
Potential popularity of the topic has nothing to do with whether the
group is properly named, appropriate for Big 8, etc. These two
issues, unrelated to each other, should not be part of the same
question.
Sufficient popularity of the topic is appropriate for a poll.
Technical correctness should be decided by people with interest in Big
8 but no interest in whether or not a particular group is created.
This decision could be made by the nan mods, a committee appointed by
them, or an elected committee, but should not be intermingled with a
poll about the popularity of the topic.
edward ohare
<mol...@canisius.edu> wrote:
>On 19 Sep 2005 14:12:23 -0700, mcc...@medieval.org (Todd Michel McComb) wrote in
><dgn9jn$nd0$1...@agricola.medieval.org>:
>
>> ... One alternative is
>>to publish addresses exactly as people use them to post, with
>>whatever munging they use. We had already discussed doing something
>>like this.
>
>That's the beauty of using the experimental groups.
Experimental groups are incompatible with the concept of management.
Mangers who do whatever they want, whenever they want, without having
to consider the costs of failure, aren't managers very long.
NAN Team
announcement, lower voting totals have indicated to us that there
is a need to lower the bar for group creation. In fact, we originally
instituted a +50 rule in place of the old +100 rule.
However, subsequent discussion raises a more basic question: Does
the process work at all? It is one thing to change +100 to +50,
but is this a pointless cosmetic change to a system which has
outlived its usefulness?
We ask that question frankly, and solicit any and all ideas for
alternatives. We are willing to consider radical change. We are
willing to consider non-radical change. We are willing to consider
no further changes, pending what we learn from the prospective fiat
creation(s).
We open this free-for-all discussion, a "Constitutional Convention"
if you will, for one month. No RFDs or CFVs will be accepted during
this time.
If, after one month, no consensus direction emerges, we will default
back to the current system (including the +50 change). If we have
concrete direction for a radical change by that time, we will have
to see how long it takes to implement. In that sense, the current
hold on RFDs is open-ended.
Finally, we apologize for creating any impression that we were
uninterested in people's thoughts on these matters. Especially as
we consider redoing the entire system, those thoughts become more
important than ever. On the personal level, I (Todd) always expect
people to comment, so it rarely occurs to me to specifically ask
for it. Certainly discussion is the backbone of news.groups, and
of course the purpose of Usenet in sum.
Todd McComb for NAN Team
Nicholas Fitzpatrick
end of the Usenet (film at 11) is the publication, and even the collection,
of e-mail addresses.
People just aren't willing to expose their e-mail addresses the way they
were years ago, in the pre-Spam era. I'm completely maniacal in making
sure my work address doesn't become published ... to the point I won't
even give it to professional societies that really should have it - because
I don't want to have to deal with the spam on that address (what I get
on this address is unbelievable ... but I have a good system in place
to deal with it - which I just can't implement at work ... and most people
aren't going to be running procmail like I do at home!).
I'm not offering any solutions (though I have some thoughts). But I think
this "conference" needs to start by identifying the issues, before we
figure out the solutions (or start the customary flame wars).
Nick
Russ Allbery
> Sufficient popularity of the topic is appropriate for a poll.
I don't really agree with this. I don't think that a poll provides an
adequate answer to this question, at least without a lot of work that
currently isn't happening reliably (getting the poll in front of the right
people) or that is intensely time-consuming (trying to eliminate multiple
answers by the same person).
Russ Allbery
Pilot projects and spike implementations are core tools in any type of
effective IT management.
Martin X. Moleski, SJ
> ... It is one thing to change +100 to +50,
>but is this a pointless cosmetic change to a system which has
>outlived its usefulness?
I'll say "yes" for now, but if, in the end, we go back to voting,
I won't complain.
>We ask that question frankly, and solicit any and all ideas for
>alternatives.
I made my proposal a few days ago. I think the best test
of a newsgroup's viability is to create it and see what happens.
If it draws enough traffic, let it live. If it doesn't, kill it.
> ... Finally, we apologize for creating any impression that we were
>uninterested in people's thoughts on these matters. ...
Nicely put. Thanks for all you guys do to keep the Big Eight
running.
Marty
Todd Michel McComb
Nicholas Fitzpatrick <nfitz@sentex> wrote:
>People just aren't willing to expose their e-mail addresses
Yes, it is pretty clear that that is an issue. I'm not sure how
we make polling work in light of that.
Personally, I'm thinking we don't want to do away with polling
entirely, but we also don't want to rely on it exclusively. I can
imagine a "fast track" system that then falls back to polling if
the committee has misgivings.
Todd McComb
mcc...@medieval.org
Tim Skirvin
>>People just aren't willing to expose their e-mail addresses
>Yes, it is pretty clear that that is an issue. I'm not sure how
>we make polling work in light of that.
This is the main problem I was hoping a web-based system would
solve.
>Personally, I'm thinking we don't want to do away with polling
>entirely, but we also don't want to rely on it exclusively. I can
>imagine a "fast track" system that then falls back to polling if
>the committee has misgivings.
I absolutely think that this is a good idea; and it seems that a
lot of others seem to agree. Jonathan Kamens' idea seemed to cover
this...
What kind of consensus do you want to reach on this to decide to
work with it? Do you want a general scheme? More of the specifics worked
out? Actual committee members?
- Tim Skirvin (tski...@killfile.org)
--
http://www.killfile.org/~tskirvin/ Skirv's Homepage <FISH>< <*>
http://www.ks.uiuc.edu/~tskirvin/pics/ Skirv's Pictures
Russ Allbery
> What kind of consensus do you want to reach on this to decide to
> work with it? Do you want a general scheme? More of the specifics worked
> out? Actual committee members?
Well, there's an order in which one would need to talk about those things,
but I think we need to work out the whole system before we start up the
newsgroup creation system again. Down to committee members if we're going
with a committee solution.
Jonathan Kamens
>entirely, but we also don't want to rely on it exclusively. I can
>imagine a "fast track" system that then falls back to polling if
>the committee has misgivings.
I think that polling should be one out of several tools that
the proponent of a group, the NAN mods, the advisory committee
I proposed, and the denizens of news.groups have at their
disposal when amassing data to decide whether to create a
group. However, I do not think that it should be a mandatory
part of the process for creating a group, and I don't think
it should be nearly so formalized as it is now.
Jonathan Kamens
addresses is one issue, but certainly not the only one or the biggest
one, with the current group creation process.
Many other issues have been brought up in recent discussion in
news.groups. They include:
* There has been rampant vote fraud.
* The procedures put in place to prevent vote fraud deter people from
voting.
* The success of the RFD/CFV for a group depends as much on the
quality of the proponents as it does on the merit of the group idea.
* There is no official, active process in place for removing obsolete
groups.
* The NAN mods do not have enough freedom to make necessary,
appropriate decisions outside of the normal RFD/CFV process.
* The current RFD/CFV process is a huge amount of work for the NAN
mods and votetakers, which makes it difficult to find people willing
to do the work.
* The time it takes to get a group created through the RFD/CFV process
is excessive when compared to the time required (essentially no time
at all) to create other kinds of forums (e.g., Yahoo Groups, Google
Groups).
* The RFD/CFV process is intended to be a tool for determining the
level of interest in a group, but (a) many people think it is no
longer effectively serving that purpose, (b) some others think it
was never accurately serving that purpose, and (c) if, indeed, it is
no longer doing what it's supposed to, it's may just be impossible
to make it do so given the current condition of the Usenet.
To address these concerns, I recently posted a proposal to eliminate
voting for newsgroups; see <dgdc1i$ikm$1...@jik2.kamens.brookline.ma.us>.
Others have posted their own proposals as well; some of them
complement mine and some contradict it. After reviewing those other
proposals, I still think mine is the right direction to go.
Todd Michel McComb
Tim Skirvin <tski...@killfile.org> wrote:
>This is the main problem I was hoping a web-based system would
>solve.
Yeah, I know that was your thought. I'm still unsure on the
registration system, or whatever method you'd use to keep track of
who is who, not to mention how the results could be inspected by
the public. I also have a personal resonance with John Stanley's
"Usenet should only use Usenet" argument. That's not meant to stop
further discussion of this option, however.
>What kind of consensus do you want to reach on this to decide to
>work with it?
I think we'll leave that question open right now, what with it being
Day 1 of 30. To do this right, I think I need to refrain from
forming any real conclusion until October 19, or at least a few
days before. I don't want to short-change the discussion by doing
anything but giving a few personal, off-the-top type of responses.
Todd McComb
mcc...@medieval.org
Larry Jaques
"Martin X. Moleski, SJ" <mol...@canisius.edu> quickly quoth:
>On Mon, 19 Sep 2005 11:51:47 -0700, newgroup...@isc.org (NAN Team) wrote
>in <11271559...@isc.org>:
>
>> ... It is one thing to change +100 to +50,
>>but is this a pointless cosmetic change to a system which has
>>outlived its usefulness?
>
>I'll say "yes" for now, but if, in the end, we go back to voting,
>I won't complain.
IF we need voting, make the requirements a lower number OR post the
bloody RFDs and RFVs on EVERY SINGLE NEWSGROUP. I wouldn't have even
known about the c.d.mysql group vote if I hadn't happened upon the
comp.lang.php group (once again) a few weeks ago. I'd be willing to
bet that 99.9999% of folks miss most of the voting since the RFDs are
so sparsely sprinkled.
>>We ask that question frankly, and solicit any and all ideas for
>>alternatives.
>
>I made my proposal a few days ago. I think the best test
>of a newsgroup's viability is to create it and see what happens.
>If it draws enough traffic, let it live. If it doesn't, kill it.
There ya go. I really like that idea, Martin. Drop the voting
requirement and just put them up, but DO advertise that they're there.
I update my groups list maybe twice a year but would do so more often
and probably join more groups if I knew that they were being formed.
Just don't make it an ALT group. Talk about ratty, spammed groups...
<sigh>
>> ... Finally, we apologize for creating any impression that we were
>>uninterested in people's thoughts on these matters. ...
>
>Nicely put. Thanks for all you guys do to keep the Big Eight
>running.
Yeah, thanks, guys!
So, did the new comp.databases.mysql group make it or not? there are 3
threads on the subject, all differing in stated outcome. (Cute trick.)
If not, let's get together a comp.lang.php.sql group.
P.S: What's with all the empty postgresql groups?!?
=====================================================================
-=Everything in Moderation,=- NoteSHADES(tm) glare guards
-=including moderation.=- http://www.diversify.com
=====================================================================
Todd Michel McComb
Jonathan Kamens <j...@kamens.brookline.ma.us> wrote:
>After reviewing those other proposals, I still think mine is the
>right direction to go.
I just want to lay out some facts and experiences when it comes to
volunteer committees. Please just consider this a bit of input
into the discussion.
It's hard to get a set of people to all have time to discuss and
conclude on something. The way we do it right now is that any one
person can post an RFD/CFV/Result, and then the others can come
back later and say "Hey, wait a minute!" if that seems prudent.
The expectation is that there isn't a lot of free time, and to have
a real "fast track" system, decisions need to be made promptly.
Now, lately, we've had to work with some real decisions, leading
to where we are right now. That's OK, and I think everyone is a
little more eager to put some time aside on this, because it's
something that really needs attention, and it's kind of a unique
situation. But 10 months from now or 3 years from now? I'm highly
skeptical that more than a tiny handful of people will be able to
put in a timely effort, if even that.
We had the current system such that it required very little judgement.
That was the great thing about it. It wasn't a question of me
having to go "Hmm, I'm going to need to set aside an actual chunk
of time to ponder the merits of this," it was "OK, boom, here you
go." That is the sort of work I was able to do promptly. And, let
me just state this pompously and upfront, I did it a lot more
promptly than anyone has in years. If I have to think "Hmm, should
this group be created?" that can be a tough question, and I know
how that can go. Now multiply it times the however-many-people who
are supposed to be involved, and I have a certain vision of how
that's going to go.
This is me talking from a lot of years on a lot of all-volunteer
working committees, so please do consider it very seriously.
Todd McComb
mcc...@medieval.org
Mean Green Dancing Machine
>entirely, but we also don't want to rely on it exclusively. I can
>imagine a "fast track" system that then falls back to polling if
>the committee has misgivings.
How do you imagine this being similar to and different from the
mechanisms used by uk.*? That should make this more concrete. Any
reason not to simply swipe the uk.* system wholesale?
--
--- Aahz <*> (Copyright 2005 by aa...@pobox.com)
Hugs and backrubs -- I break Rule 6 http://rule6.info/
Androgynous poly kinky vanilla queer het Pythonista
"It's 106 miles to Chicago. We have a full tank of gas, a half-pack of
cigarettes, it's dark, and we're wearing sunglasses." "Hit it."
Todd Michel McComb
>mechanisms used by uk.*? That should make this more concrete.
>Any reason not to simply swipe the uk.* system wholesale?
I'm not familiar enough to say. Maybe that's the way to go, though.
Todd McComb
mcc...@medieval.org
Jonathan Kamens
fraud any better than the current system. I also don't see
how it can work without requiring people at some point to give
an email address, so it also won't address people's reluctance
to give out their addresses, although it would probably be
possible to eliminate the need to *post* the addresses, munged
or not, at the end of the vote.
Having said that, I'd also like to say that although I was an
adherent of the "Usenet should only use Usenet" philosophy for
quite a while, I've pretty much abandoned it by the wayside.
To my mind, the reality is that the Usenet is no longer just
the Usenet, it's part of the Internet / World Wide Web /
whatever you want to call it. The number of people with
access to the Usenet nowadays who do not have the ability to
use the World Wide Web is small enough that I really don't
think disenfranchising them is a real issue.
Jonathan Kamens
I, too, have spent a lot of years working on a lot of
all-volunteer committees. I continue to do so, albeit not on
Usenet-related committees, and I've experienced many times
over the difficulty of getting the people on a committee to
commit time to the work. So I know where you're coming from.
However, I'm not convinced that this is a deal-breaker for my
proposed structure of governance.
We're just not seeing a lot of proposals for new groups
nowadays. By my count, there were at most 16 serious
proposals in the last year. It doesn't seem terribly
overwhelming to ask active committee members to spend a bit of
time thinking about that many proposals over the course of a
year.
I don't think you actually need "more than a tiny handful of
people" actively involved to make my proposed system work.
Those people will discuss. Those peole will make decisions.
Their decisions will be mostly reasonable. If there are any
that are egregiously unreasonable, they will be challenged by
the advisory committee, and at that point it'll become clear
to the "lurkers" that there's something out of the ordinary
that they need to spend some time thinking about.
If, on the entire Usenet, we won't be able to find a small
handful of reasonable people, agreeable to each other, who are
willing to take a stint at helping run things, then I'd say
that the Usenet is no longer a viable medium and we should
just pack up and go home. I don't think that's the case.
For the record, I'd volunteer to be on either the active
committee or the advisory committee. I wouldn't do it
forever, but I'd take my shift.
Tim Skirvin
>I don't see how a Web-based voting system can prevent vote
>fraud any better than the current system.
I don't know if it really can either; but the current system used
by the UVV is getting pretty good at it, and a web-based system wouldn't
make things any worse.
>I also don't see how it can work without requiring people at some point
>to give an email address, so it also won't address people's reluctance
>to give out their addresses, although it would probably be possible to
>eliminate the need to *post* the addresses, munged or not, at the end of
>the vote.
That was my intention. The email addresses would never be
published, though they may be available to some administrators/auditors.
- Tim Skirvin (tski...@killfile.org)
--
http://www.killfile.org/~tskirvin/ Skirv's Homepage <FISH>< <*>
http://www.killfile.org/~tskirvin/cows/ Skirv's Cows
Todd Michel McComb
Jonathan Kamens <j...@kamens.brookline.ma.us> wrote:
>I'm not convinced that this is a deal-breaker for my proposed
>structure of governance.
Oh, I'm not either. I'm not convinced of most anything right now,
by design.
>If, on the entire Usenet, we won't be able to find a small handful
>of reasonable people, agreeable to each other, who are willing to
>take a stint at helping run things, then I'd say that the Usenet
>is no longer a viable medium and we should just pack up and go home.
>I don't think that's the case.
Well, I don't want to equate the viability of the creation system
with the viability of Usenet. It's easy for me to observe that
many people are getting something good from Usenet in 2005.
Todd McComb
mcc...@medieval.org
Todd Michel McComb
Tim Skirvin <tski...@killfile.org> wrote:
>The email addresses would never be published, though they may be
>available to some administrators/auditors.
That's the tricky bit, it seems to me. Telling people their addresses
"probably" won't leak is not the easiest sell. One alternative is
to publish addresses exactly as people use them to post, with
whatever munging they use. We had already discussed doing something
like this.
Todd McComb
mcc...@medieval.org
Todd Michel McComb
saur <sa...@nyc.rr.com> wrote:
>Two weeks is all the process should take.
I'm not against your idea, but I do want to point out that there
is no guarantee that whoever volunteers to verify these sorts of
facts is going to have the spare time whenever a proponent happens
to drop by.
It's fine for Jonathan to say "it only needs to happen 16 times in
a year," but if those 16 times aren't happening when a random
outsider wants them to happen, there's going to be some unrest, and
you won't have your fast track.
>But if you want a data-driven system
I like the phrase, and what it implies. It's certainly a hot
buzzword in education right now, and mostly for the best.
Todd McComb
mcc...@medieval.org
edward ohare
wrote:
>edward ohare <edward...@nospam.yahoo.com.invalid> writes:
>
>> Sufficient popularity of the topic is appropriate for a poll.
>
>I don't really agree with this. I don't think that a poll provides an
>adequate answer to this question, at least without a lot of work that
>currently isn't happening reliably (getting the poll in front of the right
>people) or that is intensely time-consuming (trying to eliminate multiple
>answers by the same person).
So, do you think that determining whether technical flaws exist is
appropriate for a poll?
edward ohare
wrote:
>edward ohare <edward...@nospam.yahoo.com.invalid> writes:
>> "Martin X. Moleski, SJ" <mol...@canisius.edu> wrote:
>
>>> That's the beauty of using the experimental groups.
>
>> Experimental groups are incompatible with the concept of management.
>> Mangers who do whatever they want, whenever they want, without having to
>> consider the costs of failure, aren't managers very long.
>
>Pilot projects and spike implementations are core tools in any type of
>effective IT management.
Martin, however, suggests that each group created be considered an
experiment. Wholey unnecessary to consider each group creation, one
after another, as an experiment. It could be assumed, over time, that
something would be learned from these experiments that could be
applied to later events?
Russ Allbery
> Russ Allbery <r...@stanford.edu> wrote:
>> I don't really agree with this. I don't think that a poll provides an
>> adequate answer to this question, at least without a lot of work that
>> currently isn't happening reliably (getting the poll in front of the
>> right people) or that is intensely time-consuming (trying to eliminate
>> multiple answers by the same person).
> So, do you think that determining whether technical flaws exist is
> appropriate for a poll?
There aren't a lot of flaws in a newsgroup that can really be called
"technical" (in the sense that software couldn't handle them). Maybe
infeasible moderation policies or something. Most of the flaws that
people categorize under that heading are really policy issues (reposting
messages to groups to which they were not originally posted, moderating
talk.* groups, that sort of thing).
I think a poll has pluses and minuses for addressing that sort of thing.
It does give one a pretty good read on how many people actually care,
which I find frequently interesting and useful information that's often
hard to come by in a typical newsgroup discussion dominated by a few loud
voices.
Russ Allbery
"Experimental" is probably not as pedantically accurate for this technique
as another term like "provisional" or "tentative."
Mark Dodel
wrote:
-> As stated in our "Voting Threshold Adjusted for Group Passage"
-> announcement, lower voting totals have indicated to us that there
-> is a need to lower the bar for group creation. In fact, we originally
-> instituted a +50 rule in place of the old +100 rule.
->
-> However, subsequent discussion raises a more basic question: Does
-> the process work at all? It is one thing to change +100 to +50,
-> but is this a pointless cosmetic change to a system which has
-> outlived its usefulness?
->
-> We ask that question frankly, and solicit any and all ideas for
-> alternatives. We are willing to consider radical change. We are
-> willing to consider non-radical change. We are willing to consider
-> no further changes, pending what we learn from the prospective fiat
-> creation(s).
->
-> We open this free-for-all discussion, a "Constitutional Convention"
-> if you will, for one month. No RFDs or CFVs will be accepted during
-> this time.
->
-> If, after one month, no consensus direction emerges, we will default
-> back to the current system (including the +50 change). If we have
-> concrete direction for a radical change by that time, we will have
-> to see how long it takes to implement. In that sense, the current
-> hold on RFDs is open-ended.
->
-> Finally, we apologize for creating any impression that we were
-> uninterested in people's thoughts on these matters. Especially as
-> we consider redoing the entire system, those thoughts become more
-> important than ever. On the personal level, I (Todd) always expect
-> people to comment, so it rarely occurs to me to specifically ask
-> for it. Certainly discussion is the backbone of news.groups, and
-> of course the purpose of Usenet in sum.
->
-> Todd McComb for NAN Team
Todd,
I think the fact that you considered a change in the threshold from
+100 to +50 a show that you do listen to people. As to a change to a
new method for deciding new group creation, I don't see what else
would work better hen a vote of interested parties. The few
suggestions I have seen strike me as elitist (only allowing people who
post regularly to news.groups, or some other way to limit who can
voice an opinion). That said I do look forward to seeing what people
propose as alternatives.
If someone can come up with a reasonable way to identify fake votes
that would be great. I can't think of a way to validate someones
identity though, other then to make sure no more than one vote comes
from an IP address. But even that isn't completely foolproof, as
there is a troll I know of who seems to have access to posting from
multiple sources.
Mark
--
From the eComStation of Mark Dodel
http://www.os2voice.org
Warpstock 2005, Hershey, PA, Oct 6-9, 2005 - http://www.warpstock.org
Martin X. Moleski, SJ
<edward...@nospam.yahoo.com.invalid> wrote in
<n9iui159phf7kai5p...@4ax.com>:
>Experimental groups are incompatible with the concept of management.
>Mangers who do whatever they want, whenever they want, without having
>to consider the costs of failure, aren't managers very long.
Accepting your assertion for the sake of argument, the costs of
using a new newsgroup to determine whether the newsgroup
can attract a readership is cheaper and more effective (in
my way of thinking) than running a vote of any sort.
The test is directly related to the question at hand: will this
newsgroup be popular? [See ru's rule.]
Voting only indirectly--very indirectly!--correlates to usage.
The group I have helped to moderate since 1998 had
around 240 Yes votes. The number of people who have
posted to the group in any one year is in the neighborhood
of a couple of dozen. Some of the Yes voters, of course,
may be lurkers, but it's posts that keep the newsgroup
alive, not silent readers.
The cost of running the test is low: pipe the headers
through a filter to give some idea of how good the
flow is. I don't know how to do this, but it is not hard
coding for folks who code for a living. The filter could
count cross-posts, numbers of individuals posting to
the group, and frequency of traffic. Some of the big
news servers run the NNTP stream through anti-spam
filters. I see very little garbage coming out of Newsguy,
and I think Supernews cleans up its feed, too. I don't
know whether there are any off-the-shelf filters for that.
There was some web site that used to give stats on
newsgroup usage. I think it gave counts of the most
frequent contributors to the group. It might also have
had some measure of "longest thread" and average
posts per day. It was a little spooky to know that someone
was watching with those questons in mind, but that's
life in a public forum.
I grant you that setting up the flow meter is going to be
expensive for somebody. But once it's in place, it will
pay back the investment for years to come.
The cost of the current system is the work it takes to
guard against vote fraud. So far as I know, that's not
trivial. As many have suggested, it may not be much of
a measure, either, of the value of the group to its potential
audience. In my view, that's a high cost with a marginal
return.
Marty
Mark Dodel
wrote:
-> j...@kamens.brookline.ma.us (Jonathan Kamens) writes:
->
-> >I don't see how a Web-based voting system can prevent vote
-> >fraud any better than the current system.
->
-> I don't know if it really can either; but the current system used
-> by the UVV is getting pretty good at it, and a web-based system wouldn't
-> make things any worse.
->
-> >I also don't see how it can work without requiring people at some point
-> >to give an email address, so it also won't address people's reluctance
-> >to give out their addresses, although it would probably be possible to
-> >eliminate the need to *post* the addresses, munged or not, at the end of
-> >the vote.
->
-> That was my intention. The email addresses would never be
-> published, though they may be available to some administrators/auditors.
->
-> - Tim Skirvin (tski...@killfile.org)
I don't see how email addresses identify anyone. With the free email
sites people can have hundreds of email addresses. With a web based
voting an IP address can be checked and validated against a list of IP
addresses that have already voted. Yes people can get a new IP
address via DHCP or just go to a different computer so even that isn't
foolproof.
Brian Mailman
>
> Here's a basic outline of the data the Proponent would have to present here:
Sounds similar to what the regs in alt.config present.
B/
JDa™©
here to look for something totally unrelated to this.
I use to be a Usenet admin for the State of Georgia from 1994 until
1999 when we stopped having Usenet altogether. I always just ended
up following everyone else lead in the creation/deletion of news
groups. We only carried selected comp, sci, k12, news, and rec groups.
Before I retired from the state, we canned Usenet because of all the
spam and pron that was being delievered, and we did not feel that the
public would enjoy knowing that state computers had access to that
kind of data.
I believe that one of the problems is the lack of user support for
Usenet altogether. My meaning is that there are now so may other
avenues for topical information, that Usenet has almost become a
lost cause.
I for one would find it informative to know some of the demographys
for Usenet as they are today, compared to what there were in the
90's. Knowing that information could help provide the current
crop of Usenet admins (*you*) with a better understand of were to
take this subject.
Just a observation from the past!.
JDa
NAN Team wrote:
> As stated in our "Voting Threshold Adjusted for Group Passage"
> announcement, lower voting totals have indicated to us that there
> is a need to lower the bar for group creation. In fact, we originally
> instituted a +50 rule in place of the old +100 rule.
> >
> Todd McComb for NAN Team
--
^__.__ X Domain & Zone Contact (-_-) for domain 'State.Ga.US.' __.__
( ___ )------------------------- 'o' ------------- & Ga.Gov ----( ___ )
<| / | James D. Almand (JDa) - (JDA-ARIN) or (AJ829-ORG) | \ |>
<| / | X State employee, commonly referred to as ** RETIRED ** | \ |>
<| / | State of Ga - GTA/IR Data Networking, Unix Systems | \ |>
(_____)---------------------------------------------------------(_____)
JDa™©
Ken Arromdee
This would be something like "if this proposal has been delayed an unreasonable
length of time, it is considered passed". The NAN moderator should then send
a control message without further deliberation, and if different people send
out the control messages and make the decisions, the one who sends out the
control messages is authorized to do so immediately.
This isn't for general delays, but for when specific proposals are delayed
much more than others. We'd need to hash out a definition; something like
"this proposal has been delayed for three months longer than the time it took
for a newer proposal to be decided".
The purpose isn't really to pass delayed proposals, though that's a side
effect; it's to discourage the Cabal from delaying proposals in the first
place. (Of course the Cabal could always ignore the rule, but if they were
going to ignore rules they'd never bother with a pocket veto anyway, they'd
just reject the proposal outright.)
--
Ken Arromdee / arromdee_AT_rahul.net / http://www.rahul.net/arromdee
"You know, you blow up one sun and suddenly everyone expects you to walk
on water." --Samantha Carter, Stargate SG-1
Adam H. Kerman
> * Similarly, we have very few paths of communication to news
> administrators, the vast majority of whom do not read news.groups.
> We don't know what's working for them and what isn't working for
> them. I've tried to contact several key administrators in the past
> and get their input, but this ideally would be an ongoing thing
> somehow.
Hierarchy administrators are performing work on behalf of News
administrators. Would it not make sense if at least the Big 8
administrators were appointed by them?
Be nice if confidence could be expressed by running checkgroups. The ones
not running checkgroups should explain to the other News administrators
that either they refuse to do so, intending to create any group they damn
well please (and dropping out of the group of news admins who appoint the
hierarchy admins), or they think it's flawed and express no confidence and
suggest replacements.
Adam H. Kerman
>j...@kamens.brookline.ma.us (Jonathan Kamens) writes:
>>I also don't see how it can work without requiring people at some point
>>to give an email address, so it also won't address people's reluctance
>>to give out their addresses, although it would probably be possible to
>>eliminate the need to *post* the addresses, munged or not, at the end of
>>the vote.
> That was my intention. The email addresses would never be
>published, though they may be available to some administrators/auditors.
What about the proponent?
Russ Allbery
> Hierarchy administrators are performing work on behalf of News
> administrators. Would it not make sense if at least the Big 8
> administrators were appointed by them?
Sure, if you could get them to do that, but from what I've heard that's
the last thing that most news administrators want to get into doing (for
varius reasons, including uncomfortableness about how that interacts with
their commercial business models).
> Be nice if confidence could be expressed by running checkgroups. The
> ones not running checkgroups should explain to the other News
> administrators that either they refuse to do so, intending to create any
> group they damn well please (and dropping out of the group of news
> admins who appoint the hierarchy admins), or they think it's flawed and
> express no confidence and suggest replacements.
Yup, it that would be very nice. I just don't know how to get there from
here.
Per Abrahamsen
monarchy instead.
One of the ISP is the formal owner/maintainer of the hierarchy (they
bought the company that created it, and the other ISP's respect their
group list as official.
The ISP don't really want to get actively involved, so they appointed
a group of volunteers (a steering committee) who had made a good work
back in the democracy to decide which groups to create and remove.
The group has been self-supplying since then.
The steering committee does not act pro-actively. The RFD process
runs as usual in the Danish equivalent of news.groups. After the RFD
period is over, the proposer send in the final proposal to the
committee. The committee then consider the arguments for and against
raised in the group (but no arguments raised other places, in that
they are kind of like a jury), and come to a decision. Usually for or
against the proposal, but occationally they call for a non-binding
vote if they want some quantitive data to supplement the arguments
raised. The vote takes whatever form is best suited to supply the
committee with the information they need.
The main argument for and against is traffic or lack of traffic.
Sometimes the committee issues a "call for traffic" where the
proponents get some period in the closest appropriate existing group
to demonstrate that interesting discussions on the subject can occur.
In general though, the recommendation is that people try to start the
discussion on the subject *before* issuing the RFD.
The main principal advantage and disadvantage (depending on viewpoint)
of this system is that arguments count for more than opinions. The
main practical advantage is that it less bureaucratic. The main
practical disadvantage is that many good people feel they don't need
to get involved, since the steering committee is running the shop
well. But the SC needs quality input in order to run the shop well.
Esa A E Peuha
Here's an idea: news admins who want show to the world that they do
process checkgroups can make their servers' active lists available so
that anyone can compare them against the active file at isc.org; I would
be willing to run software that does such comparsion automatically.
If most news admins can't be bothered to do even that much of passive
endorsement for the hierarchy administration (whatever way that is being
done), there doesn't seem to be much point doing it at all.
--
Esa Peuha
student of mathematics at the University of Helsinki
http://www.helsinki.fi/~peuha/
Martin X. Moleski, SJ
<87aci81...@windlord.stanford.edu>:
>> Martin, however, suggests that each group created be considered an
>> experiment. Wholly unnecessary to consider each group creation, one
>> after another, as an experiment. It could be assumed, over time, that
>> something would be learned from these experiments that could be applied
>> to later events?
>"Experimental" is probably not as pedantically accurate for this technique
>as another term like "provisional" or "tentative."
I'm not wedded to the name. I didn't intend any offense to those
who think they think like scientists.
My suggestion is that the best way to find out whether a group will
"work" is to run the group.
With each new group there is a hypothesis to be tested. The
proponents argue that the group will succeed if it is created.
I say the way to find out whether their hypothesis is true is to
run the most appropriate experiment: create the group and
see what happens.
From the success of one new group, nothing much can be
inferred about the creation of another new group--the next
new group has to stand or fall on its own merits.
There would have to be some kind of death penalty that prevents re-opening
a group that fails its trial period--perhaps burial of the topic for a year
or so before the group or a clone of it could be re-opened.
A friend of mine who earns his living from engineering says, "One test
is worth a thousand expert opinions."
Marty
Martin X. Moleski, SJ
<86pll1s...@sirppi.helsinki.fi>:
>If most news admins can't be bothered to do even that much of passive
>endorsement for the hierarchy administration (whatever way that is being
>done), there doesn't seem to be much point doing it at all.
Some will follow.
That's enough (by my standards).
If admins aren't using the Big Eight lists, then they won't
be adding new newsgroups, either.
I think the mods should focus on making the best list
possible, then let the chips fall where they may.
Marty
Marcel Beaudoin
news:Pine.LNX.4.61.05...@sj.puvarg.pbz:
> Supposedly, all Usenet users (at least the ones who speak English)
> should watch n.a.n for a proposed newsgroup on a topic of interest. I
> suppose if they register at your proposed Web site, they could glance
> at it instead.
I would hazard a guess and say that the overwhelming majority of folks
(approaching 100%) currently using usenet do not even know that NAN exists.
--
Marcel
http://mudbunny.blogspot.com/
Jonathan Kamens
>I don't see how email addresses identify anyone. With the free email
>sites people can have hundreds of email addresses.
My understanding is that this is one of the reason that voting
rules were tightened up recently.
>With a web based
>voting an IP address can be checked and validated against a list of IP
>addresses that have already voted.
This is already being done with the current system -- the IP
address of the person voting is pulled from the headers of
the vote email.
I don't know how the current system deals with people voting
through proxy firewalls. That is, the issue isn't merely
people voting multiple times with different email addresses
from different IP addresses; the issue is also that separate
people may want to legitimately vote separately but show up as
coming from the same IP address.
Jonathan Kamens
One of the problems we've been discussing is that there are no
good proponents, i.e., proponents willing to do the work
necessary to get a group passed under the current system.
The amount of work you're suggesting that proponents need to
do seems like substantially more, not less, than they need to
do to be successful. That suggests to me that there will be
even fewer proponents willing to do the work, which means
even fewer newsgroups created.
The question I led off this posting with was slightly
tongue-in-cheek. I agree that proponents should be expected
to provide some hard data to support the need for a new group.
But (a) it shouldn't be be anywhere near as much data as you
suggested, (b) what data is expected should depend on the
specific circumstances of each proposal.
For example:
* The example data you listed assumes that the purpose of the
proposal is to create a forum on the Usenet for discussing
something that's already being discussed elsewhere in a forum
that is no longer working. Neither of those assumptions will
be true in every case.
* I don't see why a proposal needs to come with a "technical
host" or a "general host".
Also, I want to emphasize that one of the most important
aspects of my proposal is that the committee decides what the
process is and changes it as needed. We can discuss ideas
now for what the process might be, but I don't think we
should be discussing the imposition of a standard process on
the committee -- what we're discussing is just fodder for
them to consider when making their decisions.
Jonathan Kamens
I think Marcel's statement above is far closer to reality
than Adam's.
Jonathan Kamens
8. Thanks for posting this information!
Jonathan Kamens
>The few
>suggestions I have seen strike me as elitist (only allowing people who
>post regularly to news.groups, or some other way to limit who can
>voice an opinion).
Anyone's allowed to voice an opinion under my proposal, and
of course the committee would be expected to take those
opinions into account when making their decisions.
But you're right that in the end, the decision would be made
by the active committee, subject to challenge only by the
people on the advisory committee. The reason for that
limitation is not to be elitist, but rather to prevent the
process from being sabotaged by kooks. That is, if we allowed
anyone to formally challenge decisions of the active
committee, then it's pretty much guaranteed that every
decision would be challenged, and then we'd have to go through
the entire challenge process for every decision, which would
waste people's time and make people unwilling to serve on
either committee.
The people on the active and advisory committees would be
expected to read news.announce.newgroups and news.groups.
They would also be expected to serve as a "filter" for
challenges from people not on the committees. That is, if
someone reasonable who isn't on the committees comes along
and challenges a decision for a legitimate reason, then all
they need to do is convince one person on the committees that
the challenge is legitimate enough to proceed, and then that
one person will formally raise the challenge.
If someone is unable to convince a single person on the
committees that their challenge is legitimate, then the odds
are that it isn't.
Chris Barnes
> However, subsequent discussion raises a more basic question: Does
>
> We ask that question frankly, and solicit any and all ideas for
First, let me commend you guys for taking this step. Being willing to
change one's paradigm is *the hardest* mental step for a person to take.
Secondly, I have a question: I assume that the groups which are either
in progress now (RFD or in the CFV stage) or were just completed (such
as the mysql group) will be allowed to continue?
--
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
Chris Barnes AOL IM: CNBarnes
ch...@txbarnes.com Yahoo IM: chrisnbarnes
You always have freedom of choice, but you never have freedom of
consequence.
Chris Barnes
> of a newsgroup's viability is to create it and see what happens.
> If it draws enough traffic, let it live. If it doesn't, kill it.
(More to Todd/Nan team)
Does Martin need to repost his proposal under this thread (which means
I'll need to repost my reply)?
Chris Barnes
> IF we need voting, make the requirements a lower number OR post the
> bloody RFDs and RFVs on EVERY SINGLE NEWSGROUP.
I *completely* disagree with this. I don't want things like that
cluttering up my newsgroups (I'm the moderator of rec.hunting and
rec.hunting.dogs).
But, that does not mean I do not understand WHY you are suggesting that.
I would suggest that the prohibition against xposting RFD or CFV to
other newsgroups and/or email lists, and/or web bbs' be removed. The
odds that the proposer of a group and/or the votetaker know of all the
places where interested parties might be lurking is virtually nil.
Chris Barnes
> I think that polling should be one out of several tools that
> the proponent of a group, the NAN mods, the advisory committee
> I proposed, and the denizens of news.groups have at their
> disposal when amassing data to decide whether to create a
> group. However, I do not think that it should be a mandatory
> part of the process for creating a group, and I don't think
> it should be nearly so formalized as it is now.
I like your idea about using voting as A method of decision making.
However, if voting is to be used, it really does need to have a
formalized process (and frankly, I very much like the process in place
now). Having a process has many, many advantages and (assuming privacy
concerns are addressed) is not really that troublesome.
Todd Michel McComb
Chris Barnes <ch...@txbarnes.com> wrote:
>Does Martin need to repost his proposal under this thread (which
>means I'll need to repost my reply)?
No, of course not. Everyone saw it.
Todd McComb
mcc...@medieval.org
Todd Michel McComb
Jonathan Kamens <j...@kamens.brookline.ma.us> wrote:
>The amount of work you're suggesting that proponents need to do
>seems like substantially more,
Well, the committee could gather some of this information. Presumably
they'd be more efficient at doing so, for one thing.
I don't know about people who will be on the committee, but talking
for myself, I like having concrete reasons. If there are some
formulas that can be developed to establish concrete reasons, all
the better, and that's something that can be tried & tested over
time. Doing things on the fly, over and over, can be really painful.
Todd McComb
mcc...@medieval.org
Todd Michel McComb
Chris Barnes <ch...@txbarnes.com> wrote:
>I have a question: I assume that the groups which are either in
>progress now (RFD or in the CFV stage) or were just completed (such
>as the mysql group) will be allowed to continue?
As per other announcements, the mysql and postgresql situations will
be decided shortly (tomorrow, I guess). I don't expect anything else
to proceed to a vote. If one of the seemingly discarded proposals
comes back asking for a vote, they'll just have to wait.
Todd McComb
mcc...@medieval.org
Tim Skirvin
>For the record, I'd volunteer to be on either the active committee or the
>advisory committee. I wouldn't do it forever, but I'd take my shift.
I'd also volunteer, if it seems appropriate.
- Tim Skirvin (tski...@killfile.org)
--
http://www.killfile.org/~tskirvin/ Skirv's Homepage <FISH>< <*>
http://www.killfile.org/~tskirvin/resume.html Skirv's Resume
Mark Dodel
-> Mark Dodel <madode...@ptd.net> wrote:
->
-> > As to a change to a new method for deciding new group creation, I don't
-> > see what else would work better hen a vote of interested parties.
->
-> The "interested parties" aren't voting anymore. That *is* the problem. The
-> cause(s) are debatable, but at this point in time it is highly unlikely that
-> any voting system will entice the "interested parties" into participating.
The problem there is that the existing procedure made it difficult to
get the word out to people who would be interested. And it is just
the nature of the beast that people will only bother to respond if its
a minimum of fuss to do so.
Mark
--
From the eComStation of Mark Dodel
http://www.os2voice.org
Warpstock 2005, Hershey, PA, Oct 6-9, 2005 - http://www.warpstock.org
Tim Skirvin
>>The email addresses would never be published, though they may be
>>available to some administrators/auditors.
>That's the tricky bit, it seems to me. Telling people their addresses
>"probably" won't leak is not the easiest sell.
But telling them "we will not post these addresses" is easier.
Users are used to sites being hacked; as long as we put in a reasonable
effort to make this site secure, I don't think they'll be too upset.
As for what would be published - names, and perhaps links to
entries on the voting pages describing their history. No email address,
unless the user explicity allows it.
(Hell, if you wanted to encourage voting, make it into a social
network of its own, and let people post whatever information about
themselves that they want... Make it into a self-managed "who's who of
Usenet". But perhaps that's a bit ambitious.)
- Tim Skirvin (tski...@killfile.org)
--
http://www.killfile.org/~tskirvin/ Skirv's Homepage <FISH>< <*>
http://www.killfile.org/~tskirvin/moderate/ Skirv's Moderation
Tim Skirvin
>It's fine for Jonathan to say "it only needs to happen 16 times in
>a year," but if those 16 times aren't happening when a random
>outsider wants them to happen, there's going to be some unrest, and
>you won't have your fast track.
...but if you've got five members on your fast track committee,
and three of them are actually around, then you can still probably make
some decisions. Or you can toss in a couple of alternates, for when some
of these guys know they'll be gone.
- Tim Skirvin (tski...@killfile.org)
--
http://www.killfile.org/~tskirvin/ Skirv's Homepage <FISH>< <*>
http://tskirvin.livejournal.com/ Skirv's LiveJournal
Tim Skirvin
>> That was my intention. The email addresses would never be
>>published, though they may be available to some administrators/auditors.
>What about the proponent?
I'm not sure I trust proponents.
- Tim Skirvin (tski...@killfile.org)
--
http://www.killfile.org/~tskirvin/ Skirv's Homepage <FISH>< <*>
http://www.ks.uiuc.edu/~tskirvin/pics/ Skirv's Pictures
Jonathan Kamens
>I like your idea about using voting as A method of decision making.
>However, if voting is to be used, it really does need to have a
>formalized process
I disagree. A formalized process makes sense only if (a) it
is necessary to make a large flow of information manageable or
(b) the vote is binding, to prevent accusations of fraud.
There haven't been enough people voting recently to make the
information flow large enough to be unmanageable by hand, and
in the proposal we're discussing, the vote wouldn't be
binding. It therefore doesn't require a formalized process.
Frankly, if it should ever become necessary to conduct a vote
under my proposed system, it seems to me that it would be
unlikely to consist of anything more than, "Hey, send me
email and let me know what you think, and I'll collate and
post the results."
Much of the system we have now was designed to give the kooks
no place on which to hang (baseless) accusations of fraud,
rigged votes, etc. In contrast, my proposal assumes
implicitly that the people doing the work (active committee)
and the people advising them (advisory committee) are
trustworthy, so mechanisms for protecting them against
accusations of untrustworthiness are pointless make-work.
And yes, by the way, there is protection in my proposal
against inappropriate behavior -- an advisory committee which
can challenge the decisions of the active committee, said
advisory committee being large enough to ensure that it can't
be subverted.
I think it's high time we stopped wasting time and effort
maintaining a system which caters to the lowest common
denominator.
>(and frankly, I very much like the process in place
>now).
Why? As many people have indicated, the process in place now
does a very poor job of measuring the level of interest in a
group, and running the current process is very
labor-intensive. What's to like?
edward ohare
<mol...@canisius.edu> wrote:
>I'm not wedded to the name. I didn't intend any offense to those
>who think they think like scientists.
I think the word "experiment" perfectly describes the process you have
proposed.
>My suggestion is that the best way to find out whether a group will
>"work" is to run the group.
I agree.
>There would have to be some kind of death penalty that prevents re-opening
>a group that fails its trial period--perhaps burial of the topic for a year
>or so before the group or a clone of it could be re-opened.
The product of the nan mods is a list of viable groups. Its one thing
to experiment in private, more or less, say Brian's hosting of a group
only on his server. Its quite another to experiment with what is
supposed to be the finished product.
>A friend of mine who earns his living from engineering says, "One test
>is worth a thousand expert opinions."
Yes, fine, but where is the test going on? What you've proposed is a
General Motors style of testing, where they sell the cars and
re-engineer them based on customer complaints.
The primary consumer of the nan mod's product are the people who run
servers. I can't imagine they'd be too thilled about adding 20 groups
every quarter and removing 15 failed experimental groups at the same
time. It seems some might take the approach of "well, we'll consider
adding these groups after we see what traffic is like", which, of
course, if enough do that, guarantees there won't be sufficient
traffic.
Unless a lot of servers create a group immediately upon receipt of a
Big 8 control, then the ability to create successful groups degrades.
I think that adopting your method will result in degraded ability to
create successful groups.
edward ohare
wrote:
>There aren't a lot of flaws in a newsgroup that can really be called
>"technical" (in the sense that software couldn't handle them).
OK, then I've misused the term, from your perspective.
>Maybe
>infeasible moderation policies or something. Most of the flaws that
>people categorize under that heading are really policy issues (reposting
>messages to groups to which they were not originally posted, moderating
>talk.* groups, that sort of thing).
Bad group names, bad placement...
>I think a poll has pluses and minuses for addressing that sort of thing.
>It does give one a pretty good read on how many people actually care,
>which I find frequently interesting and useful information that's often
>hard to come by in a typical newsgroup discussion dominated by a few loud
>voices.
I plead guilty on the last. <G>
Walter Roberson
Ken Arromdee <arro...@green.rahul.net> wrote:
:Since we're suggesting ideas, I'll suggest pocket veto protection.
You're suggesting we newsgroup nerds should have pocket protectors??
;-)
--
I am spammed, therefore I am.
Adam H. Kerman
>"Adam H. Kerman" <a...@chinet.com> writes:
>>> That was my intention. The email addresses would never be
>>>published, though they may be available to some administrators/auditors.
>>What about the proponent?
> I'm not sure I trust proponents.
I'm sorry to hear you say that, Tim. If an interested party had no right
to review the information gathered, let alone audit it, he would have
legitimate reason to believe that it was kept from him because it doesn't
support the reported result.
Adam H. Kerman
>Marcel Beaudoin <mbea...@scintrextrace.com> writes:
>>>Supposedly, all Usenet users (at least the ones who speak English)
>>>should watch n.a.n for a proposed newsgroup on a topic of interest. I
>>>suppose if they register at your proposed Web site, they could glance
>>>at it instead.
>>I would hazard a guess and say that the overwhelming majority of folks
>>(approaching 100%) currently using usenet do not even know that NAN exists.
>I think Marcel's statement above is far closer to reality than Adam's.
He didn't quote my comment from that message, "In reality, most Usenet
users are complacent in these matters."
Graham Drabble
news:dgp6fj$j8d$1...@news.tamu.edu:
> I would suggest that the prohibition against xposting RFD or
> CFV to other newsgroups and/or email lists, and/or web bbs' be
> removed.
There is no prohibitation against reposting RFDs.
--
Graham Drabble
Want help with an RFD?
Try the RFDMaker: http://users.ox.ac.uk/~sjoh1646/rfd/
Graham Drabble
news:11271559...@isc.org:
> However, subsequent discussion raises a more basic question: Does
> the process work at all? It is one thing to change +100 to +50,
> but is this a pointless cosmetic change to a system which has
> outlived its usefulness?
I think the process does still work although there may be other ways to
do it. One think I would like to see is a change from YES-NO > 50 to
YES > 50 (or any other number you choose). The requirement for a
majority is built in to the 2/3rds yes requirement, the +50 part is
designed to guage interest and therefore shouldn't have anything to do
with opposition.
uk.* has a fast-track process where, if the RFD is uncontroversial the
proponent can apply to the committee for a fast track. If the committee
agrees it is uncontroversial and that it has a reasonable amount of
support then it is published. People then get 5 days to object. If more
than 6 people object or any 1 person objects with a reason the
committee decides is "well founded" then the fast track fails. The
proponent can then either go for another RFD or call a vote.
--
Graham Drabble
http://www.drabble.me.uk/
Graham Drabble
news:tskirvin.20050920154240$62...@cairo.ks.uiuc.edu:
> ...but if you've got five members on your fast track
> committee, and three of them are actually around, then you can still
> probably make some decisions. Or you can toss in a couple of
> alternates, for when some of these guys know they'll be gone.
The uk.* committee tends to work on a majority of those expressing an
opinion rather than a majority of those entitled to (although in
practice the consensus tends to be obvious and no formal counting is
done).
Marcel Beaudoin
news:Pine.LNX.4.61.05...@sj.puvarg.pbz:
I interpreted your comment to mean that they knew about NAN and/or the
intricacies of newsgroup creation and just didn't care or don't mind others
taking care of it for them.
My apologies for the misinterpretation.
--
Marcel
http://mudbunny.blogspot.com/
Adam H. Kerman
>Russ Allbery <r...@stanford.edu> writes:
>>Adam H Kerman <a...@chinet.com> writes:
>>>Be nice if confidence could be expressed by running checkgroups. The
>>>ones not running checkgroups should explain to the other News
>>>administrators that either they refuse to do so, intending to create any
>>>group they damn well please (and dropping out of the group of news
>>>admins who appoint the hierarchy admins), or they think it's flawed and
>>>express no confidence and suggest replacements.
>>Yup, it that would be very nice. I just don't know how to get there from
>>here.
>Here's an idea: news admins who want show to the world that they do
>process checkgroups can make their servers' active lists available so
>that anyone can compare them against the active file at isc.org; I would
>be willing to run software that does such comparsion automatically.
Wouldn't it be better to compare it directly to a recent checkgroups as
there could have been a processing glitch when creating the sample active
file at ftp.isc.org?
In any event, I can see how something like that could work. Some sort of
central registry of institutional News servers: Commercial, university, or
any News server supporting a large user base of employees or people in a
community, or even friends I suppose (for some large number).
I don't think I'd give a lot more weight to servers with the largest of
all user bases, as Usenet's flavor and variety comes from users on smaller
servers. Perhaps a bit more weight would be fair, but not precisely
proportional to the number of users.
In any event, I'm sure a workable mechanism could be designed. The problem
is how to get News administrators to participate in indirect decision
making for the Big 8.
>If most news admins can't be bothered to do even that much of passive
>endorsement for the hierarchy administration (whatever way that is being
>done), there doesn't seem to be much point doing it at all.
That's my feeling too.
Russ Allbery
On the other hand, I'm with Tim; as a voter, I *definitely* don't trust
all proponents. Now, I personally don't consider my e-mail address to be
in any way restricted information, but if I were someone who did, this
would be an issue for me.
--
Russ Allbery (r...@stanford.edu) <http://www.eyrie.org/~eagle/>
Russ Allbery
> Wouldn't it be better to compare it directly to a recent checkgroups as
> there could have been a processing glitch when creating the sample
> active file at ftp.isc.org?
I must admit that I don't currently check this on an automated basis, but
I should. ftp.isc.org really should have exactly the same list of
newsgroups as a current checkgroups, faster than the checkgroups is
reposted.
> In any event, I'm sure a workable mechanism could be designed. The
> problem is how to get News administrators to participate in indirect
> decision making for the Big 8.
Exactly.
Tim Skirvin
>>>> That was my intention. The email addresses would never be
>>>>published, though they may be available to some administrators/auditors.
>>>What about the proponent?
>> I'm not sure I trust proponents.
>I'm sorry to hear you say that, Tim.
What, you think we should trust them? If somebody were to propose
rec.comp.soc.i-hate-adam.talk, should they be entitled to see all of the
email addresses of those that voted "no"?
>If an interested party had no right to review the information gathered,
>let alone audit it, he would have legitimate reason to believe that it
>was kept from him because it doesn't support the reported result.
I'm all about independent reviewers. Perhaps the Usenet Volunteer
Auditors could help with this?
- Tim Skirvin (tski...@killfile.org)
--
http://www.killfile.org/~tskirvin/ Skirv's Homepage <FISH>< <*>
http://www.killfile.org/~tskirvin/archives/ Skirv's Archives
Jonathan Kamens
>In article <dgp9gs$fv8$1...@jik2.kamens.brookline.ma.us>,
>Jonathan Kamens <j...@kamens.brookline.ma.us> wrote:
>>I disagree. A formalized process makes sense only if (a) it
>>is necessary to make a large flow of information manageable or
>>(b) the vote is binding, to prevent accusations of fraud.
>
>
>Otherwise it'll be easy to give a group less consideration because someone
>on the NAN team doesn't like the subject and is less inclined to think the
>group is worthy to be created without any real reason to believe so.
People who can't be trusted to consider proposals fairly
shouldn't be on the committees.
There is a challenge process to address the possibility of a
proposal being torpedoed by inappropriate bias.
Adam H. Kerman
>On 20 Sep 2005 12:55:20 +0300, Esa A E Peuha <esa....@helsinki.fi> wrote:
>>If most news admins can't be bothered to do even that much of passive
>>endorsement for the hierarchy administration (whatever way that is being
>>done), there doesn't seem to be much point doing it at all.
>Some will follow.
>That's enough (by my standards).
>If admins aren't using the Big Eight lists, then they won't
>be adding new newsgroups, either.
That's not a reasonable assumption.
I think some News servers started life by grabbing an active list from
another server, and adding new groups on a selected basis. It's possible
they process all newgroups, or newgroups they've reviewed for usefulness
to their user base, or they honor user requests (perhaps not checking to
see if the group is recognized). rmgroups may or may not be ignored.
Perhaps the problem is that tale hadn't issued checkgroups. When Russ
began to issue them, certain News administrators didn't change their
habits to process them from time to time.
>I think the mods should focus on making the best list
>possible, then let the chips fall where they may.
That'll happen regardless as hierarchy administration is not central
administration of Usenet.
I'm stating the opinion that News administrators should participate
indirectly in administration of these hierarchies (demonstrating their
confidence by processing checkgroups or making the case for replacement of
the hierarchy administrators if they've fallen out of favor). Quite
frankly, it would have to be Todd and Russ that recruit their
participation, as they are the only ones with standing to do so.
The News administrators to recruit would be the ones who support a
large user base (no idea what threshold that should be) and create a full
set of Big 8 newsgroups. They should run checkgroups or give a
satisfactory explanation as to why it isn't run, to be considered on a
case by case basis. They should not generally be willing to create
newsgroups not recognized by the hierarchy administrators.
Now, if Russ thinks the commercial ones won't be recruitable due to
unwillingness to reveal business practices to competitors, all I can say
to that is, we won't know until they are asked.
Does anyone not understand how big a mess there would be if the Big 8 went
to a temporary newsgroup system AND checkgroups weren't regularly run?
Temporary newsgroups would seem to require that checkgroups be run
monthly. If we had no confidence that this was being done on a substantial
number of sites, it would be a mistake to send such newgroups.
Adam H. Kerman
>Who said we wanted a data-driven system?
Yo.
>One of the problems we've been discussing is that there are no
>good proponents, i.e., proponents willing to do the work
>necessary to get a group passed under the current system.
>The amount of work you're suggesting that proponents need to
>do seems like substantially more, not less, than they need to
>do to be successful. That suggests to me that there will be
>even fewer proponents willing to do the work, which means
>even fewer newsgroups created.
Gosh. Seems to me there would be a true incentive to do the work, as it
would be known in advance that the goal of getting a newgroup issued was
within reach.
Who the hell needs proponents unwilling to discuss the topic? I don't know
why anyone should care about their suggestions for what topics need new
groups. New groups should be created on behalf of those participating.
>The question I led off this posting with was slightly
>tongue-in-cheek. I agree that proponents should be expected
>to provide some hard data to support the need for a new group.
>But (a) it shouldn't be be anywhere near as much data as you
>suggested, (b) what data is expected should depend on the
>specific circumstances of each proposal.
>For example:
>* The example data you listed assumes that the purpose of the
>proposal is to create a forum on the Usenet for discussing
>something that's already being discussed elsewhere in a forum
>that is no longer working. Neither of those assumptions will
>be true in every case.
>* I don't see why a proposal needs to come with a "technical
>host" or a "general host".
Shouldn't someone come forward who is highly motivated to get the group up
and running?
>Also, I want to emphasize that one of the most important
>aspects of my proposal is that the committee decides what the
>process is and changes it as needed. We can discuss ideas
>now for what the process might be, but I don't think we
>should be discussing the imposition of a standard process on
>the committee -- what we're discussing is just fodder for
>them to consider when making their decisions.
The committee will not act in lieu of a proponent. They won't take charge
of a new group and see that discussion gets started.
Adam H. Kerman
>Exactly.
If there were a concensus on a workable mechanism, would you be willing to
recruit them? You are known worldwide; it would have to be you. I'm sure
people would be willing to assist in putting together a contact list and
with some of the follow through.
If a commercial site refused to participate, I suppose their users who are
also following this process could be induced to apply a bit of pressure by
threatening to exercise consumer choice in a free market.
Todd Michel McComb
Jonathan Kamens <j...@kamens.brookline.ma.us> wrote:
>my proposed system,
Maybe the proposed committee members will have a better sense of
some of these proposals than I do, but I can't envision polling
being all that uncommon. You can phrase it more casually, but
you'll still want honest information... as per another subthread,
*some* kind of data. At least that's my opinion.
Todd McComb
mcc...@medieval.org
Adam H. Kerman
>"Adam H. Kerman" <a...@chinet.com> writes:
>>>>> That was my intention. The email addresses would never be
>>>>>published, though they may be available to some administrators/auditors.
>>>>What about the proponent?
>>> I'm not sure I trust proponents.
>>I'm sorry to hear you say that, Tim.
> What, you think we should trust them? If somebody were to propose
>rec.comp.soc.i-hate-adam.talk, should they be entitled to see all of the
>email addresses of those that voted "no"?
Would they be voting "no" because it's an improperly named flame group, or
only because the flames wouldn't be seen by a wide enough Usenet audience
if not posted in the groups in which I usually post my dopey opinions?
>>If an interested party had no right to review the information gathered,
>>let alone audit it, he would have legitimate reason to believe that it
>>was kept from him because it doesn't support the reported result.
> I'm all about independent reviewers. Perhaps the Usenet Volunteer
>Auditors could help with this?
I understand that there's no practical way to keep proponents with bad
motive out of the system. But the ones with good motive would still have
legitimate reason to question a result. It's not reasonable to expect them
to trust a third party not of their choosing to be an arbitrator. They
wouldn't even have an advocate in the contest.
Todd Michel McComb
Ken Arromdee <arro...@green.rahul.net> wrote:
>Any idea which involves people being there should have a fallback
>for the situation where the people aren't there (or refuse to
>respond).
We actually have some formal procedures right now. We've never
needed to turn to that, because we've never had a real disagreement,
but we have them just in case. One aspect is that if someone hasn't
responded to previous decisions, we don't even need to wait for
them on the current decision. There's also a time limit in there.
We also ask people to specifically state when they will be away,
which I always do (I'm the one who is away a lot).
But, seriously, any system of committees needs to be designed against
the Path of Most Sloth, i.e. unless there's a big brouhaha, the
person who cares the most is going to be performing that task by
himself most of the time. That's the way things actually work.
Todd McComb
mcc...@medieval.org
Walter Roberson
And if the newsgroup had to do with the discussion of issues
pertaining to what some people might consider a dangerous cult?
Or if the newsgroup discussed matters that might present
difficult legal challenges, such as music or video sharing, or
satellite descrambling?
--
Chocolate is "more than a food but less than a drug" -- RJ Huxtable
Russ Allbery
> Quite frankly, it would have to be Todd and Russ that recruit their
> participation, as they are the only ones with standing to do so.
I've made a few efforts in that direction, although I can't say that I've
really gone all out, and the results weren't promising.