Advertising and Retro-Mod
Richard E. Depew
that it is a solution for a non-existant problem.
My previous answer to this has been that some newsgroups face
problems that are severe enough to cause at least some persons in that
newsgroup to suggest making the group moderated.
I'd like to present a second justification in this note...
preparing for anticipated problems as Usenet grows.
One of the problems that will arise is that as Usenet grows it
becomes more attractive as a place to advertise. We have recently
read (or joined) a discussion on the place of advertising on the net,
with the majority opinion seeming to be that it was OK in its place
(specific newsgroups or replying to inquiries), but that general
unsolicited advertising was not OK. A minority opinion was that
since it wasn't illegal, it couldn't be stopped.
Let's compare the response to an onslaught of advertising in
a hypothetical democratic newsgroup with a typical anarchic newsgroup...
call them "comp.compression.research" and "news.future" or "ccr" and
"nf".
Situation: John Palmer lands a multi-million dollar contract
to promote "Preparation H" on the network. He begins by changing the
name of his site (once again) to "tygra.preparation-h.com" and the
name of his BBS from "cat-talk" to "preparation-h talk". He then
sends a commercial to comp.new-prod, where Chip rejects it. JP then
fires up a cron job to walk his active file once a day, posting a
"PREPARATION-H TALK IS BACK!!!!!!" message to every valid newsgroup
that he can find. He does this though uunet, who appreciate the
increased traffic.
day 1
The retromoderators in comp.compression.research suspect that
this ad has little to do with compression research, and MOD the first
one that comes in, warning that if such off-charter postings continue,
they will take stronger action. The readers of news.future respond
with dozens of flames, predictions of the imminent death of Usenet,
proposals to let readers score commercials. uunet's profits rise.
One user cancels the ad and is immediately hounded off the net for
interfering with First Amendement Rights.
day 2
The retromoderators in comp.compression.research use the SKP
function, telling the readership that the message has little to do
with the newsgroup. The readers of news.future respond to the second
ad with dozens more flames, directed not at the ad, but at each other.
One poster from a Canadian site complains that the ad is not posted in
a French-only edition. This results in demands for versions in twenty
different languages... and several hilarious spoofs are posted.
day 3
The retromoderators in comp.compression.research use the ROT13
function, telling the readership that the message is just a repeat of
the first two day's messages. The readers of news.future respond to
the third ad with by exchanging examples of the Path: headers from all
three articles in an effort to find the source of the ad. However, JP
has botched his configuration again, so they all seem to come from
uunet. The traffic through uunet rises. The WSJ reports that sales
of perparation-h have soared.
day 4
The retromoderators in comp.compression.research use the CAN
function, vowing to cancel any more that come in. The readers of
news.future post dozens more flames, three new uses for preparation H,
and two claims that "this is the best of all possible nets". Jiro
sendsys bombs JB, setting off the UDD, and the world-wide web melts down...
all because news.future didn't accept retromoderation in time. :-)
Dick
--
Richard E. Depew, Rootstown, OH r...@uhura.neoucom.edu (work)
``The freshmen bring a little knowledge in and the seniors take none
out, so it accumulates through the years.'' -- A. Lawrence Lowell,
President of Harvard
Wes Morgan
r...@uhura.neoucom.edu (Richard E. Depew) wrote:
> Let's compare the response to an onslaught of advertising in
>a hypothetical democratic newsgroup with a typical anarchic newsgroup...
Situation: [paraphrased to avoid insult -- Wes] Johnny Advert
sets up a posting robot to inflict one advert each
day on our hypothetical newsgroup.
>day 1
> The retromoderators [...] MOD the first one that comes in...
>day 2
> The retromoderators [...] use the SKP function...
>day 3
> The retromoderators [...] use the ROT13 function...
>day 4
> The retromoderators [...] use the CAN function...
Days 1-3 achieve nothing but even-more-increased traffic as the MOD,
SKP, and ROT13 control messages (including, presumably, multiple
copies from multiple retro-moderators) fly around the net to catch up
to the original postings. Uunet and other pay-for-feed providers are,
indeed, greatly pleased; in fact, they offer the retro-moderators com-
missions on the increased fee revenues and a "keep up the good work"
commendation letter.
Let's see here.....let's say that the newsgroup has 4 dedicated
retro-moderators, each of whom takes each action you mention. This
would not be unusual, since retro-moderators probably won't see the
action taken by their counterparts elsewhere.....and they'll be quick
on the trigger, since they have to move fast to catch up to the post.
- One original message per day * 4 days 4
- One MOD control msg * 4 moderators 4
- One SKP control msg * 4 moderators 4
- One ROT control msg * 4 moderators 4
- One CAN control msg * 4 moderators 4
-------------------------------------------------------
Total 20
The retro-moderators finally get around to a cancel on day 4.
After four days and anywhere from 8 to 20 postings (depending
on who gets which control messages when; I suspect that some
poor site at the convergence of these feeds might have to pro-
cess all 20) we've saved our hypothetical newsgroup from the
blandishments of (the now properly chastened) Johnny Advert.
Some sites have shunted the control messages around long enough
to figure out that they're redundant, but they're still paying
to do so, while other non-direct-paying sites just clog up their
NNTP feeds.......
I won't even consider the possibility of disagreement among
retro-moderators....."Oh, no, that isn't deserving of a ROT;
I'll just supersede his ROT with a SKP." Yeesh........there's
also the possibility of rogue retro-moderators undoing all of
this just for the heck of it (as Dick said, the software it-
self is rather trivial.)........double yeesh.
Most users have seen 3 conflicting versions of the posting (SKP,
ROT and MOD). Users at sites close to the poster's site have
probably seen the 4 originals too, so they might have seen 7
copies.
Going just a bit further, let's suppose that someone *replies*
to Johnny Advert's message(s). Will the retro-moderators go
ahead and kill off those messages too, or might we see another
MOD/SKP/ROT/CAN cycle? Geez......
For comparison purposes:
Traditional moderation -- message 1 never gets off the
virtual runway. Impact on sites
(other than that of moderator) is nil.
Readers w/o killfiles -- Examine headers, realize that it's
an advert, hit 'n' key. Impact on
site: 4 messages + enough CPU time
to process 'n' key.
Readers with killfiles -- readers who do not wish to see
adverts take care of it after
message 1 or 2. Impact on sites is
4 messages + enough CPU time to pro-
cess killfile entry.
Retro-moderation -- Readers may see multiple copies of same
message, may have to unROT/unSKP to figure
out if he's already seen it. Impact on
sites: 4 originals + anywhere from 4-16
control messages + CPU/feed time required
to propagate 4-16 control messages.
Refresh my memory -- this is supposed to be an *improvement*?
--Wes
--
Wes Morgan - University of Kentucky - Lexington, Kentucky USA
mor...@engr.uky.edu || morgan@ukcc || ...!ukma!ukecc!morgan
Mailing list for AT&T StarServer E/S admins -- starserve...@engr.uky.edu
GCS/E/MU d--- -p+ c++ l+ m* s++/++ !g w+ t+ r
Cousin It
| Let's compare the response to an onslaught of advertising in
|a hypothetical democratic newsgroup with a typical anarchic newsgroup...
|call them "comp.compression.research" and "news.future" or "ccr" and
|"nf".
| Situation: John Palmer lands a multi-million dollar contract
|to promote "Preparation H" on the network. He begins by changing the
|name of his site (once again) to "tygra.preparation-h.com" and the
|name of his BBS from "cat-talk" to "preparation-h talk". He then
|sends a commercial to comp.new-prod, where Chip rejects it. JP then
|fires up a cron job to walk his active file once a day, posting a
|"PREPARATION-H TALK IS BACK!!!!!!" message to every valid newsgroup
|that he can find. He does this though uunet, who appreciate the
|increased traffic.
Sorry. All you are saying is that John is being highly
annoying, but what else is new. I see nothing illegal, and without
more info, nothing off charter. You can't delete something for being
annoying.
Oh, and your little scenario is humorous, but has little
substance.
Brandon Magee c...@cs.brandeis.edu
Business Manager/Acting General Manager Live Music Engineer
WBRS 100 FM, Waltham Ma. WB...@binah.cc.brandeis.edu
All Genre Music Programming Free Live Music-4 times weekly
Richard E. Depew
Wes, you have disappointed me. After all our discussion about
Message-IDs you still have not understood that no matter how many
retroactive moderators act on a particular off-charter post, each
site will see only one retromod version of that article. Johnny, on
the other hand, will get (if his email address is working) a copy of
the charter each time a retroactive moderator acts.
>The retro-moderators finally get around to a cancel on day 4.
>After four days and anywhere from 8 to 20 postings (depending
>on who gets which control messages when; I suspect that some
>poor site at the convergence of these feeds might have to pro-
>cess all 20) we've saved our hypothetical newsgroup from the
>blandishments of (the now properly chastened) Johnny Advert.
>Some sites have shunted the control messages around long enough
>to figure out that they're redundant, but they're still paying
>to do so, while other non-direct-paying sites just clog up their
>NNTP feeds.......
When the message ID of a retromod version of an article is offered
via NNTP to a site that already has a retromod version of the same article
from a different retroactive moderator, it will tell the feed site
that it already has the article. The only extra traffic you can blame
on multiple moderators is the message-IDs.
>I won't even consider the possibility of disagreement among
>retro-moderators....."Oh, no, that isn't deserving of a ROT;
>I'll just supersede his ROT with a SKP." Yeesh........there's
>also the possibility of rogue retro-moderators undoing all of
>this just for the heck of it (as Dick said, the software it-
>self is rather trivial.)........double yeesh.
You are stretching, Wes. You might as well invoke the Mutlu/
Argic Bot while you are at it. :-)
>Most users have seen 3 conflicting versions of the posting (SKP,
>ROT and MOD). Users at sites close to the poster's site have
>probably seen the 4 originals too, so they might have seen 7
>copies.
Don't forget that they will also the "CAN" announcement. That
makes up to 8. This is still better than the flame fest that will
be going on in news.future. :-)
>Going just a bit further, let's suppose that someone *replies*
>to Johnny Advert's message(s). Will the retro-moderators go
>ahead and kill off those messages too, or might we see another
>MOD/SKP/ROT/CAN cycle? Geez......
I'd recommend "CAN". :-)
>For comparison purposes:
>
> Traditional moderation -- message 1 never gets off the
> virtual runway. Impact on sites
> (other than that of moderator) is nil.
Agreed.
> Readers w/o killfiles -- Examine headers, realize that it's
> an advert, hit 'n' key. Impact on
> site: 4 messages + enough CPU time
> to process 'n' key.
But this will go on forever (or until net melt-down).
> Readers with killfiles -- readers who do not wish to see
> adverts take care of it after
> message 1 or 2. Impact on sites is
> 4 messages + enough CPU time to pro-
> cess killfile entry.
Likewise... this will not deter Johnny.
> Retro-moderation -- Readers may see multiple copies of same
> message, may have to unROT/unSKP to figure
> out if he's already seen it. Impact on
> sites: 4 originals + anywhere from 4-16
> control messages + CPU/feed time required
> to propagate 4-16 control messages.
There is some overhead, true, but there is also deterrence.
>Refresh my memory -- this is supposed to be an *improvement*?
If it eventually discourages Johnny from posting his ads to the
newsgroup, it will have been worth the extra traffic, yes. If not,
it may be time to set up ARMM again.
Dick
--
Richard E. Depew, Munroe Falls, OH r...@redpoll.mrfs.oh.us (home)
``The liberty of the individual must be thus far limited; he must not
make himself a nuisance to other people.'' -- John Stuart Mill
Richard E. Depew
>r...@uhura.neoucom.edu (Richard E. Depew) writes:
>
>| Let's compare the response to an onslaught of advertising in
>|a hypothetical democratic newsgroup with a typical anarchic newsgroup...
>|call them "comp.compression.research" and "news.future" or "ccr" and
>|"nf".
>
>| Situation: John Palmer lands a multi-million dollar contract
>|to promote "Preparation H" on the network. He begins by changing the
>|name of his site (once again) to "tygra.preparation-h.com" and the
>|name of his BBS from "cat-talk" to "preparation-h talk". He then
>|sends a commercial to comp.new-prod, where Chip rejects it. JP then
>|fires up a cron job to walk his active file once a day, posting a
>|"PREPARATION-H TALK IS BACK!!!!!!" message to every valid newsgroup
>|that he can find. He does this though uunet, who appreciate the
>|increased traffic.
>
> Sorry. All you are saying is that John is being highly
>annoying, but what else is new. I see nothing illegal, and without
>more info, nothing off charter. You can't delete something for being
>annoying.
The off-charter bit comes from the generally (but not
universally) accepted idea that commercial advertising should be
restricted to the "biz" hierarchy or to newsgroups that explicitly
invite such posts. I'd quote from the newuser documents on this
topic, but they seem to have expired here.
> Oh, and your little scenario is humorous, but has little
>substance.
I'm pleased you see the humor, and I'm glad that this is not yet
a real situation. It is, however, possible that we'll face something
like this in the future. Retro-Mod will be ready for it. :-)
Dick
--
Edmund Schweppe
>mor...@engr.uky.edu (Wes Morgan) writes:
>>
>>r...@uhura.neoucom.edu (Richard E. Depew) wrote:
>>> Let's compare the response to an onslaught of advertising in
>>>a hypothetical democratic newsgroup with a typical anarchic newsgroup...
>> Situation: [paraphrased to avoid insult -- Wes] Johnny Advert
>> sets up a posting robot to inflict one advert each
>> day on our hypothetical newsgroup.
Note that in Dick's scenario, Johnny Advert was smart enough to write a
roboposter, but not smart enough to figure out that he could cut down his
uunet bills by crossposting his adverts to multiple groups. Even the
"MAKE MONEY FAST" posters usually figure that one out beforehand. I will
assume that the retromoderators will jump directly to EJEction in this
case, since that is the only function Dick has described which deals with
crossposts.
If, in fact, the adverts were crossposted to, say, five unmoderated and
one retromoderated group, the analysis becomes much simpler. Each day,
the readers of the unmoderated groups get to see at *least* two copies of
the post; one from Johnny Advert and one from the retromoderator(s) who
are busy EJEcting the adverts as fast as they can see them. The readers
of the retromoderated group will see all the EJEction messages, and some
percentage of the daily adverts, depending on the timing between the
arrival of the ejections and the reading of the news by the individual
users.
And if Johnny Advert crossposts to *two* retromod and four unmoderated
groups, the fun really starts! If retromoderator #1 and retromoderator #2
both post their EJE's at roughly the same time, then some percentage of
the Net will see the advert being reinjected into group #1, and some
percentage will see the advert being reinjected into group #2. The
retromoderators, though, will *never* see the reinjections since both of
their machines already have something with the forged Message-ID. So,
pretty soon, the readers of the groups will recognize that their
retromoderators are *not* doing their job and the flamefest will erupt
once again. Meanwhile, the unmoderated groups are getting *two* copies a
day of the advert.
On the other hand, if retromoderator #1's ejection arrives at
retromoderator #2's site before retromoderator #2 ejects the original
advert, then the situation is not quite so bad for the retromod groups
(that day - the timings are likely to differ day to day). Retrogroup #1
sees the original and #1's ejection, while retrogroup #2 sees the
original, #1's reinjection, and #2's ejection. The poor SOBs in the
unmoderated groups, meanwhile are up to *three* copies per day of Johnny's
advertising.
>>Going just a bit further, let's suppose that someone *replies*
>>to Johnny Advert's message(s). Will the retro-moderators go
>>ahead and kill off those messages too, or might we see another
>>MOD/SKP/ROT/CAN cycle? Geez......
> I'd recommend "CAN". :-)
Even for the cross-post case? What if one or more of the groups
cross-posted to had "democratically voted" to *allow* and *encourage*
advertising?
>>For comparison purposes:
>> Retro-moderation -- Readers may see multiple copies of same
> There is some overhead, true, but there is also deterrence.
I ask again, HOW?!?
If a user does not respond to social pressures, there is no valid reason
to assume that he or she will respond to *ineffective* cancellations. I
emphasize ineffective because, as Brad Templeton has pointed out, sysops
on single-machine systems *have* effective means to stop threads available
to them.
Further, there is absolutely nothing in the world to prevent Johnny Advert
from adding a quick "repost-the-original-when-a-cancel-comes-in" piece of
code to his news software. And if he is in fact pulling a feed directly
from uunet with a quick turnaround time, I can absolutely guarantee that
his code can post them faster than human retromoderators can read his
postings, decide that they are "bad", and cancel or eject them.
Meanwhile, the net traffic continues to soar.
>>Refresh my memory -- this is supposed to be an *improvement*?
> If it eventually discourages Johnny from posting his ads to the
>newsgroup, it will have been worth the extra traffic, yes. If not,
>it may be time to set up ARMM again.
Why do I get the feeling that the rehabilitation of ARMM is the real
reason behind the entire "retroactive moderation / newsgroup democracy"
concept?
--
__________________________________________________________________________
Edmund Schweppe - schw...@bumetb.bu.edu, schw...@acs.bu.edu,
(or for faster replies) eschweppe%drcoa1...@vax3.drc.com
All standard disclaimers (also datclaimers and deotherclaimers) apply.
Wes Morgan
>> - One original message per day * 4 days 4
>> - One MOD control msg * 4 moderators 4
>> - One SKP control msg * 4 moderators 4
>> - One ROT control msg * 4 moderators 4
>> - One CAN control msg * 4 moderators 4
>> -------------------------------------------------------
>> Total 20
>
> Wes, you have disappointed me. After all our discussion about
>Message-IDs you still have not understood that no matter how many
>retroactive moderators act on a particular off-charter post, each
>site will see only one retromod version of that article. Johnny, on
>the other hand, will get (if his email address is working) a copy of
>the charter each time a retroactive moderator acts.
Minor correction here, Dick; the *users* will only see on retromod
version of the article. The *feeds* will still have to process all
of them, even if it's only long enough to decide that they already
have it. It's still overhead, and there will still be up to 20
control messages flying around the net (if not reaching each par-
ticular site).
> When the message ID of a retromod version of an article is offered
>via NNTP to a site that already has a retromod version of the same article
>from a different retroactive moderator, it will tell the feed site
>that it already has the article. The only extra traffic you can blame
>on multiple moderators is the message-IDs.
So there is no provision for disagreement among retro-moderators;
whoever acts first has the definitive say.........doesn't sound
very democratic.
> You are stretching, Wes. You might as well invoke the Mutlu/
>Argic Bot while you are at it. :-)
Why not? You invoked ARMM, which is essentially the same thing.
You believe, to this day, that your intent was noble; surely you
do not discount the existence of others with similarly "lofty"
goals?
> Don't forget that they will also the "CAN" announcement. That
>makes up to 8. This is still better than the flame fest that will
>be going on in news.future. :-)
User education, not ad hoc retro-moderation, is the solution.
Since retro-moderation affects with is made available to third
parties, it is *not* the same thing.
>> Readers w/o killfiles -- Examine headers, realize that it's
>> an advert, hit 'n' key. Impact on
>> site: 4 messages + enough CPU time
>> to process 'n' key.
>
> But this will go on forever (or until net melt-down).
No, not quite; it will go on until sufficient consensus *across the
net* makes action possible. Don't think that such consensus is im-
possible; Robert McElwaine would disagree with you.
>>Refresh my memory -- this is supposed to be an *improvement*?
>
> If it eventually discourages Johnny from posting his ads to the
>newsgroup, it will have been worth the extra traffic, yes. If not,
>it may be time to set up ARMM again.
It will be a *long* time before ARMM should ever see the light of
day. In fact, I'd suggest that ARMM should have 'activation keys'
held by the admins of the Top 10 feed sites; only unanimous agree-
ment among that control group should *ever* reenable ARMM.
Cousin It
| Wes, you have disappointed me. After all our discussion about
|Message-IDs you still have not understood that no matter how many
|retroactive moderators act on a particular off-charter post, each
|site will see only one retromod version of that article. Johnny, on
|the other hand, will get (if his email address is working) a copy of
|the charter each time a retroactive moderator acts.
You dissapoint yourself. Each individual reader sees at most
two postings per day, however, the site itself sees all 5. Oh, and
Johnny Advert Uses elm and filters all such charter mail to /dev/null
and sees nothing.
[Wes:]
|>I won't even consider the possibility of disagreement among
|>retro-moderators....."Oh, no, that isn't deserving of a ROT;
|>I'll just supersede his ROT with a SKP." Yeesh........there's
|>also the possibility of rogue retro-moderators undoing all of
|>this just for the heck of it (as Dick said, the software it-
|>self is rather trivial.)........double yeesh.
| You are stretching, Wes. You might as well invoke the Mutlu/
|Argic Bot while you are at it. :-)
You think he is stretching. But, people disagree all the time,
and this may be a time where it will happen. [Actually, It could be
fun. One moderator thinks that something should have a ctl-l, then the
next retro-moderator sees that, thought it didn't need anything and
takes it out, ad nauseum]
| Don't forget that they will also the "CAN" announcement. That
|makes up to 8. This is still better than the flame fest that will
|be going on in news.future. :-)
No it isn't. It is much worse. Flamefests are fun.
|>Going just a bit further, let's suppose that someone *replies*
|>to Johnny Advert's message(s). Will the retro-moderators go
|>ahead and kill off those messages too, or might we see another
|>MOD/SKP/ROT/CAN cycle? Geez......
| I'd recommend "CAN". :-)
Of course you would. I suggest CAN for anything you post. :/
|> Readers w/o killfiles -- Examine headers, realize that it's
|> an advert, hit 'n' key. Impact on
|> site: 4 messages + enough CPU time
|> to process 'n' key.
| But this will go on forever (or until net melt-down).
Your point?
|> Readers with killfiles -- readers who do not wish to see
|> adverts take care of it after
|> message 1 or 2. Impact on sites is
|> 4 messages + enough CPU time to pro-
|> cess killfile entry.
| Likewise... this will not deter Johnny.
So?
|> Retro-moderation -- Readers may see multiple copies of same
|> message, may have to unROT/unSKP to figure
|> out if he's already seen it. Impact on
|> sites: 4 originals + anywhere from 4-16
|> control messages + CPU/feed time required
|> to propagate 4-16 control messages.
| There is some overhead, true, but there is also deterrence.
Johhny doesn't give a flying fig. He keeps doing it. It is not
a deterrence.
|>Refresh my memory -- this is supposed to be an *improvement*?
| If it eventually discourages Johnny from posting his ads to the
|newsgroup, it will have been worth the extra traffic, yes. If not,
|it may be time to set up ARMM again.
Thats a big IF. And... ARMM will just get you cut off.
Brandon Magee c...@cs.brandeis.edu
"We all see black and white, when it comes to someone elses fight
no one ever gets involved, apathy can never solve"
Anthrax "Indians"
Cousin It
|>r...@uhura.neoucom.edu (Richard E. Depew) writes:
|> Sorry. All you are saying is that John is being highly
|>annoying, but what else is new. I see nothing illegal, and without
|>more info, nothing off charter. You can't delete something for being
|>annoying.
| The off-charter bit comes from the generally (but not
|universally) accepted idea that commercial advertising should be
|restricted to the "biz" hierarchy or to newsgroups that explicitly
|invite such posts. I'd quote from the newuser documents on this
|topic, but they seem to have expired here.
I never knew advertising your free access usenet site had to
be put in the biz hierarchy. I guess we have to get rid of From:
headers because we are advertising our hosts. And no, the fact that
John has named it after a hemmerhoid medicine does not matter.
|> Oh, and your little scenario is humorous, but has little
|>substance.
| I'm pleased you see the humor, and I'm glad that this is not yet
|a real situation. It is, however, possible that we'll face something
|like this in the future. Retro-Mod will be ready for it. :-)
And the people opposed to retro-mod will be ready for it too.
Brandon Magee c...@cs.brandeis.edu
"The nation's glued to CNN, to watch their own creation"
-Winger "Blind Revolution Mad"
Heidi de Wet
>a hypothetical democratic newsgroup with a typical anarchic newsgroup...
>call them "comp.compression.research" and "news.future" or "ccr" and
>"nf".
>
>day 2
> The retromoderators in comp.compression.research use the SKP
>function, telling the readership that the message has little to do
>with the newsgroup. The readers of news.future respond to the second
>ad with dozens more flames, directed not at the ad, but at each other.
>One poster from a Canadian site complains that the ad is not posted in
>a French-only edition. This results in demands for versions in twenty
>different languages... and several hilarious spoofs are posted.
You've hit the nail on the head. This alone is sufficient reason
to scrap retromoderation.
--
--------- Heidi de Wet --------- he...@eleceng.uct.ac.za ---------
"Error can point the way to truth, while empty-headedness can only
lead to more empty-headedness or to a career in politics."
- Master Li, _Bridge of Birds_
Richard E. Depew
>In <CABL5...@redpoll.mrfs.oh.us>, r...@redpoll.mrfs.oh.us (Richard E. Depew) writes:
>
>>mor...@engr.uky.edu (Wes Morgan) writes:
>>>
>>>r...@uhura.neoucom.edu (Richard E. Depew) wrote:
>>>> Let's compare the response to an onslaught of advertising in
>>>>a hypothetical democratic newsgroup with a typical anarchic newsgroup...
>>> Situation: [paraphrased to avoid insult -- Wes] Johnny Advert
>>> sets up a posting robot to inflict one advert each
>>> day on our hypothetical newsgroup.
>
>Note that in Dick's scenario, Johnny Advert was smart enough to write a
>roboposter, but not smart enough to figure out that he could cut down his
>uunet bills by crossposting his adverts to multiple groups.
That is not an unreasonable scenario if you read biz.comp.services.
>Even the
>"MAKE MONEY FAST" posters usually figure that one out beforehand. I will
>assume that the retromoderators will jump directly to EJEction in this
>case, since that is the only function Dick has described which deals with
>crossposts.
I was trying to find a scenario where all the retromod functions
might be exercised. If a "Preparation H" commercial were cross-posted
to many groups, and if none were groups that this ad seemed
appropriate for (e.g. alt.tasteless), then the retromoderator may
simply "CAN" it from the beginning. That would have made the scenario
less interesting. "EJE" is to be used when the cross-posted article
may be on-charter for the other group(s) in the newsgroup list.
[...byzantine EJEs leading to multiple copies of the ad]
Edmund, you seem to be assuming that the retroactive moderators
will behave as complete idiots. Yes, an idiot retroactive moderator
can undoubtedly make things worse. My advice to the newsgroup is
"don't elect idiots as retroactive moderators." You don't give a
loaded gun to an idiot.
>>>Going just a bit further, let's suppose that someone *replies*
>>>to Johnny Advert's message(s). Will the retro-moderators go
>>>ahead and kill off those messages too, or might we see another
>>>MOD/SKP/ROT/CAN cycle? Geez......
>> I'd recommend "CAN". :-)
>
>Even for the cross-post case? What if one or more of the groups
>cross-posted to had "democratically voted" to *allow* and *encourage*
>advertising?
Well, in that case, "EJE", editing down the new newsgroup line
to include just those newsgroups that *want* this sort of post.
>>>For comparison purposes:
>>> Retro-moderation -- Readers may see multiple copies of same
>> There is some overhead, true, but there is also deterrence.
>
>I ask again, HOW?!?
By the constant reminders to the poster that these postings are
off-charter, unwanted, and canceled at the earliest opportunity.
>If a user does not respond to social pressures, there is no valid reason
>to assume that he or she will respond to *ineffective* cancellations. I
>emphasize ineffective because, as Brad Templeton has pointed out, sysops
>on single-machine systems *have* effective means to stop threads available
>to them.
Retroactive moderation is not a cure-all, only an aid. It is a
slap on the wrist, meant to get the attention of the poster so he will
reconsider his actions.
>Further, there is absolutely nothing in the world to prevent Johnny Advert
>from adding a quick "repost-the-original-when-a-cancel-comes-in" piece of
>code to his news software. And if he is in fact pulling a feed directly
>from uunet with a quick turnaround time, I can absolutely guarantee that
>his code can post them faster than human retromoderators can read his
>postings, decide that they are "bad", and cancel or eject them.
>Meanwhile, the net traffic continues to soar.
At this point it is time to admit that stronger measures are
needed. Measures similar to those currently used.
>>>Refresh my memory -- this is supposed to be an *improvement*?
>> If it eventually discourages Johnny from posting his ads to the
>>newsgroup, it will have been worth the extra traffic, yes. If not,
>>it may be time to set up ARMM again.
>
>Why do I get the feeling that the rehabilitation of ARMM is the real
>reason behind the entire "retroactive moderation / newsgroup democracy"
>concept?
There is a kernel of truth in this observation. ARMM was the
first draft of a system for defending newsgroups against certain posts.
Dick
--
Richard E. Depew, Munroe Falls, OH r...@redpoll.mrfs.oh.us (home)
``The greatest blessing of our democracy is freedom. But in the last
analysis, our only freedom is the freedom to discipline ourselves.''
-- Bernard Baruch
Heidi de Wet
>>I will
>>assume that the retromoderators will jump directly to EJEction in this
>>case, since that is the only function Dick has described which deals with
>>crossposts.
>
>to many groups, and if none were groups that this ad seemed
>appropriate for (e.g. alt.tasteless), then the retromoderator may
>simply "CAN" it from the beginning.
Let me get this right: you're advocating that a retromoderator for
group A take it upon themselves to cancel an article posted to another
group just because it was crossposted to group A? What happened
to the "will of the people"? The people in the other group haven't
voted for retromoderation, they may in fact have voted against it, and
they certainly didn't nominate that particular person as being fit to
judge what is, or is not, tasteless.
>That would have made the scenario less interesting.
It would have made the scenario quite untenable. It still does.
> Edmund, you seem to be assuming that the retroactive moderators
>will behave as complete idiots. Yes, an idiot retroactive moderator
>can undoubtedly make things worse. My advice to the newsgroup is
>"don't elect idiots as retroactive moderators." You don't give a
>loaded gun to an idiot.
My advice to you is, think up some sort of safeguard (if possible) for
the inevitable day when an idiot, or thug, gets hold of your software.
Don't suggest that the admins then step in to rescue you from the
consequences of your stupidity; that's not acceptable (they'll simply
step in now and pre-empt the whole mess.) If there is, as I suspect,
no adequate safeguard, then don't manufacture the weapons in the first
place. (Think nuclear weapons, Dick. Think about what would happen if
anyone could FTP a nuclear bomb.)
>>> I'd recommend "CAN". :-)
>>
>>Even for the cross-post case? What if one or more of the groups
>>cross-posted to had "democratically voted" to *allow* and *encourage*
>>advertising?
>
> Well, in that case, "EJE", editing down the new newsgroup line
>to include just those newsgroups that *want* this sort of post.
You can't pretend that a retromoderator has a mandate to interpret
any charter but that of the group they're moderating.
That's the problem with vigilanteism: condone it in one particular
circumstance, and it soon gets completely out of hand.
Edmund Schweppe
>schw...@bumetb.bu.edu (Edmund Schweppe) writes:
>>In <CABL5...@redpoll.mrfs.oh.us>, r...@redpoll.mrfs.oh.us (Richard E. Depew) writes:
>>>mor...@engr.uky.edu (Wes Morgan) writes:
>>>>r...@uhura.neoucom.edu (Richard E. Depew) wrote:
>>>>> Let's compare the response to an onslaught of advertising in
>>>>>a hypothetical democratic newsgroup with a typical anarchic newsgroup...
>>>> Situation: [paraphrased to avoid insult -- Wes] Johnny Advert
>>>> sets up a posting robot to inflict one advert each
>>>> day on our hypothetical newsgroup.
>>Note that in Dick's scenario, Johnny Advert was smart enough to write a
>>roboposter, but not smart enough to figure out that he could cut down his
>>uunet bills by crossposting his adverts to multiple groups.
> I was trying to find a scenario where all the retromod functions
>might be exercised.
Understood; however, it started me thinking about the implications of
crossposting in a retromod environment, and ...
>If a "Preparation H" commercial were cross-posted
>to many groups, and if none were groups that this ad seemed
>appropriate for (e.g. alt.tasteless), then the retromoderator may
>simply "CAN" it from the beginning. That would have made the scenario
>less interesting. "EJE" is to be used when the cross-posted article
>may be on-charter for the other group(s) in the newsgroup list.
WHAT?!? Where exactly does the retroactive moderator of group ABC derive
*any* authority or *any* justification whatsoever for cancelling a posting
to group DEF?
The ONLY shred of justification you have presented so far for the actions
of retroactive moderators is that the newsgroup affected has, at some
point, requested and consented to retroactive moderation by certain
specified retromoderators. Now you are claiming that retroactive
moderators have the right and the privilege to censor *ALL* groups on the
Net? Even traditionally moderated groups? Even groups which have not
voted to be retrocensored by anyone? Even groups who have *specifically*
voted *against* retromoderation?
There is ABSOLUTELY NO JUSTIFICATION for that sort of action. Remember
what happened the last time somebody arbitrarily appointed themselves the
moderator of a hierarchy?
>[...byzantine EJEs leading to multiple copies of the ad]
[ Including the possibility of a crosspost to two or more retromod groups ]
> Edmund, you seem to be assuming that the retroactive moderators
>will behave as complete idiots. Yes, an idiot retroactive moderator
>can undoubtedly make things worse. My advice to the newsgroup is
>"don't elect idiots as retroactive moderators." You don't give a
>loaded gun to an idiot.
No, I was assuming (a) that the retroactive moderator(s) of one group have
no way of knowing whether or not another group in a crosspost is
retroactively moderated, and (b) that said retromoderators will limit
their activities to the group which "elected" them. Condition (a) comes
from your insistance on not changing any current news software, so that
news servers and news readers cannot tell a retromod group from an
unmoderated group. I thought that condition (b) held from your claims
that retroactive moderation would not affect current newsgroups unless
they specifically asked for it.
Judging from your comments above, I was mistaken. Being "elected"
retroactive moderator by some minor group is a licence to kill messages
anywhere on Usenet - which is just one of the many objections that have
raised to the concept of retroactive moderation.
>>>>Going just a bit further, let's suppose that someone *replies*
>>>>to Johnny Advert's message(s). Will the retro-moderators go
>>>>ahead and kill off those messages too, or might we see another
>>>>MOD/SKP/ROT/CAN cycle? Geez......
>>> I'd recommend "CAN". :-)
>>Even for the cross-post case? What if one or more of the groups
>>cross-posted to had "democratically voted" to *allow* and *encourage*
>>advertising?
> Well, in that case, "EJE", editing down the new newsgroup line
>to include just those newsgroups that *want* this sort of post.
And just how does the retroactive moderator *know* which groups "*want*
this sort of post" and which don't? The answer is that he (or she) has no
idea whatsoever.
Even if you assume that the retromoderator is omniscient, and that by
virtue of her election as retromoderator of foo.bar.bletch has the
authority to remove unwarranted posts from foo.bar.bletch, she *still* has
absolutely no right at all to cancel something from bar.bletch.foo.
>>>>For comparison purposes:
>>>> Retro-moderation -- Readers may see multiple copies of same
>>> There is some overhead, true, but there is also deterrence.
>>I ask again, HOW?!?
> By the constant reminders to the poster that these postings are
>off-charter, unwanted, and canceled at the earliest opportunity.
Which are, as has been noted in the past, promptly filtered to /dev/null.
Or, if the roboposter is truly mischievous, forwarded back to the sender.
Remember that you are assuming that the *current* system of mailing
complaints to the "bad poster" is ineffective - else there would be no
reason to consider retroactive moderation in the first place.
>>If a user does not respond to social pressures, there is no valid reason
>>to assume that he or she will respond to *ineffective* cancellations. I
>>emphasize ineffective because, as Brad Templeton has pointed out, sysops
>>on single-machine systems *have* effective means to stop threads available
>>to them.
> Retroactive moderation is not a cure-all, only an aid. It is a
>slap on the wrist, meant to get the attention of the poster so he will
>reconsider his actions.
In other words - ineffective and not meant to be effective. It sounds
more like some sort of childish vengeance - "you put this crap on my
machine! Hah! Take that - I cancelled you! Pfffppptt!"
>>Further, there is absolutely nothing in the world to prevent Johnny Advert
>>from adding a quick "repost-the-original-when-a-cancel-comes-in" piece of
>>code to his news software. And if he is in fact pulling a feed directly
>>from uunet with a quick turnaround time, I can absolutely guarantee that
>>his code can post them faster than human retromoderators can read his
>>postings, decide that they are "bad", and cancel or eject them.
>>Meanwhile, the net traffic continues to soar.
> At this point it is time to admit that stronger measures are
>needed. Measures similar to those currently used.
Meanwhile, there has been a considerable amount of excess net pollution.
The sysadmins of the sites where the retromoderators work are getting
flooded with complaints from the members of the other groups; if the
retromoderators reinject the advertisements, those who dislike the ads
complain about crap being reposted; if the retromoderators simply cancel
the advertisements, those who like the ads complain about censorship.
*Both* sides will have legitimate grounds for complaint.
>>>>Refresh my memory -- this is supposed to be an *improvement*?
>>> If it eventually discourages Johnny from posting his ads to the
>>>newsgroup, it will have been worth the extra traffic, yes. If not,
>>>it may be time to set up ARMM again.
>>Why do I get the feeling that the rehabilitation of ARMM is the real
>>reason behind the entire "retroactive moderation / newsgroup democracy"
>>concept?
> There is a kernel of truth in this observation. ARMM was the
>first draft of a system for defending newsgroups against certain posts.
Against their will, mind you.
Francisco X DeJesus
> One complaint about my retroactive moderation proposals has been
>that it is a solution for a non-existant problem.
>
> My previous answer to this has been that some newsgroups face
>problems that are severe enough to cause at least some persons in that
>newsgroup to suggest making the group moderated.
>
> I'd like to present a second justification in this note...
>preparing for anticipated problems as Usenet grows.
>
> One of the problems that will arise is that as Usenet grows it
>becomes more attractive as a place to advertise.
[...ridiculous death-of-usenet scenario deleted...]
First off: anyone who reposts an ad for a real product over and over
again, annoying people left and right, will simply find the hard way
that this is the wrong approach to advertising. They would be dooming
their product/service/whatever to getting a bad name worldwide by
upsetting everyone repeatedly. So you see, that problem takes care of
itself rather nicely...
Now, if retro-mod was in place, what we might have would be the opposite
effect. You would be giving them free advertising... like people who rush
to protest a new "controversial" movie that offends them, and then find that
their media coverage has simply attracted more people to see the movie.
Simple reverse psychology. Plus, of course, in the process you spam the
network with increased overhead traffic and upset people even more as
you make them read the same post time and time again, in MOD, SKP, ROT,
etc versions.
[BTW Dick, I do speak from some experience here. I worked for one of the
largest ad agencies in the country for a while, and if there's one rule
they follow, it's "don't piss off the buying public"... do so intentionally
or knowing full well that that's the expected reaction, and you've got
yourself someone looking for a new job]
So to sum up: bad example, Dick. Thanks for playing. Try again.
--
Francisco X DeJesus ----- S A I C ----- dej...@c3ot.saic.com
------------------------------------------------------------------------------
* disclaimer: Opinions expressed here are mine. Typos and errors are yours *
"Duck Season!" "Rabbit Season!" "...rabbit season." "It's Duck Season! SHOOT!"
Wes Morgan
>schw...@bumetb.bu.edu (Edmund Schweppe) writes:
>>I will
>>assume that the retromoderators will jump directly to EJEction in this
>>case, since that is the only function Dick has described which deals with
>>crossposts.
>
>If a "Preparation H" commercial were cross-posted
>to many groups, and if none were groups that this ad seemed
>appropriate for (e.g. alt.tasteless), then the retromoderator may
>simply "CAN" it from the beginning.
Holy cow! I didn't notice this the first time around......I've
got to pay better attention to your subtle expansions of power,
Dick.
Do we now see a retromoderator elected for newsgroup X cancelling
a message in newsgroup Y? Tsk, tsk, tsk.....this is hardly put-
ting power in the hands of each "newsgroup community."
Even *you* can't stay within the bounds of your proposal, Dick;
how can you expect a corps of retro-moderators to do so?
--Wes
--
Wes Morgan - University of Kentucky - mor...@engr.uky.edu
Mailing list for AT&T StarServer E/S admins - starserve...@engr.uky.edu
GCS/E/MU d--- -p+ c++ l+ m* s++/++ !g w+ t+ r
gharshana-neti -- mental floss?
Ken Arromdee
>First off: anyone who reposts an ad for a real product over and over
>again, annoying people left and right, will simply find the hard way
>that this is the wrong approach to advertising. They would be dooming
>their product/service/whatever to getting a bad name worldwide by
>upsetting everyone repeatedly. So you see, that problem takes care of
>itself rather nicely...
First off: advertising departments have a slow response time. By the time
the information that people are pissed filters down to those responsible for
the ads, it could be several months.
Second: if they get enough customers, it really doesn't matter how much they
piss everyone else off. Consider telemarketers: maybe 95% of people called
by telemarketers get pissed off, but they don't care, because they get their
customers from the other 5%.
--
"On the first day after Christmas my truelove served to me... Leftover Turkey!
On the second day after Christmas my truelove served to me... Turkey Casserole
that she made from Leftover Turkey.
[days 3-4 deleted] ... Flaming Turkey Wings! ...
-- Pizza Hut commercial (and M*tlu/A*gic bait)
Ken Arromdee (arro...@jyusenkyou.cs.jhu.edu)
Cousin It
|>Even the
|>"MAKE MONEY FAST" posters usually figure that one out beforehand. I will
|>assume that the retromoderators will jump directly to EJEction in this
|>case, since that is the only function Dick has described which deals with
|>crossposts.
| I was trying to find a scenario where all the retromod functions
|might be exercised. If a "Preparation H" commercial were cross-posted
|to many groups, and if none were groups that this ad seemed
|appropriate for (e.g. alt.tasteless), then the retromoderator may
|simply "CAN" it from the beginning.
I am sorry, this is why you made the EJE function. The
moderator is hardly the conscience of the group he is moderating, lets
not make him the conscience of groups HE IS NOT MODERATING.
|That would have made the scenario
|less interesting. "EJE" is to be used when the cross-posted article
|may be on-charter for the other group(s) in the newsgroup list.
Hardly less interesting. EJE is to be used when there is a
crosspost to your group which you feel does not belong in said group.
Don't you dare tell me it doesn't belong anywhere else.
| Edmund, you seem to be assuming that the retroactive moderators
|will behave as complete idiots. Yes, an idiot retroactive moderator
|can undoubtedly make things worse. My advice to the newsgroup is
|"don't elect idiots as retroactive moderators." You don't give a
|loaded gun to an idiot.
And you don't give armms to a Richard. I'm sorry Dick, it's a
specious argument. Advice is only that, advice. One can not stop the
election of idiots, especially one who has a latent tendency towards
idiocy.
|>Even for the cross-post case? What if one or more of the groups
|>cross-posted to had "democratically voted" to *allow* and *encourage*
|>advertising?
| Well, in that case, "EJE", editing down the new newsgroup line
|to include just those newsgroups that *want* this sort of post.
Sorry, again Johny Moderator is acting as the conscience of
groups which don't want him as said conscience.
|>> There is some overhead, true, but there is also deterrence.
|>
|>I ask again, HOW?!?
| By the constant reminders to the poster that these postings are
|off-charter, unwanted, and canceled at the earliest opportunity.
Sorry, you are wrong. All that will do is enrage them. The
best way to deal with a usual idiot is to ignore them. The best way to
deal with a persistent idiot is to ignore them. The only time you want
to deal with an idiot is when the idiot wishes to take actions [real
actions, not posts] that can hurt the newsgroup, news systems, or are
otherwise harm causing.
|>If a user does not respond to social pressures, there is no valid reason
|>to assume that he or she will respond to *ineffective* cancellations. I
|>emphasize ineffective because, as Brad Templeton has pointed out, sysops
|>on single-machine systems *have* effective means to stop threads available
|>to them.
| Retroactive moderation is not a cure-all, only an aid. It is a
|slap on the wrist, meant to get the attention of the poster so he will
|reconsider his actions.
A piece of mailing stating why the post is innapropriate for
the news.group according to you is a hell of a lot better aid. And it
doesn't carry the overhead.
|>Further, there is absolutely nothing in the world to prevent Johnny Advert
|>from adding a quick "repost-the-original-when-a-cancel-comes-in" piece of
|>code to his news software. And if he is in fact pulling a feed directly
|>from uunet with a quick turnaround time, I can absolutely guarantee that
|>his code can post them faster than human retromoderators can read his
|>postings, decide that they are "bad", and cancel or eject them.
|>Meanwhile, the net traffic continues to soar.
| At this point it is time to admit that stronger measures are
|needed. Measures similar to those currently used.
And which are those? The same ones I, and others, have said
are currently innefective?
|>Why do I get the feeling that the rehabilitation of ARMM is the real
|>reason behind the entire "retroactive moderation / newsgroup democracy"
|>concept?
| There is a kernel of truth in this observation. ARMM was the
|first draft of a system for defending newsgroups against certain posts.
A horrible first draft for an equally loathable idea.
Francisco X DeJesus
>schw...@bumetb.bu.edu (Edmund Schweppe) writes:
[...]
>>Why do I get the feeling that the rehabilitation of ARMM is the real
>>reason behind the entire "retroactive moderation / newsgroup democracy"
>>concept?
>
> There is a kernel of truth in this observation. ARMM was the
>first draft of a system for defending newsgroups against certain posts.
Coming up with Bad Ideas just to give ARMM (a monumentally Bad Idea)
something to do (or as others would put it, "scrabbling to save face") is
in itself, a Bad Idea.
D. J. Bernstein
> that it is a solution for a non-existant problem.
Indeed. Dick, where's your public support? Why can't you just go away
until other people ask for your help in setting up a retro-mod group?
---Dan
Francisco X DeJesus
>In article <CAGzp...@avalon.chinalake.navy.mil> dej...@avalon.nwc.navy.mil (Francisco X DeJesus) writes:
>>First off: anyone who reposts an ad for a real product over and over
>>again, annoying people left and right, will simply find the hard way
>>that this is the wrong approach to advertising. They would be dooming
>>their product/service/whatever to getting a bad name worldwide by
>>upsetting everyone repeatedly. So you see, that problem takes care of
>>itself rather nicely...
>
>First off: advertising departments have a slow response time. By the time
>the information that people are pissed filters down to those responsible for
>the ads, it could be several months.
BZZZT! Wrong guess! Care to go for Double Jeopardy, where the scores
can really change?
See, for the most part, companies hire advertising agencies. It is they
who purchase ads, not an "advertising department". And when an ad upsets
people they generally call the TV station, network, radio station,
newspaper, magazine, whatever. These "carriers" are notorious for
passing the buck as fast as they can ("Sorry, we didn't make up the ad. We
will be sure to contact the responsible parties right away"). So it
usually comes back to the agency *very* quickly. As the agency's reputation
(which covers more than one product or brand) is on the line, quick
action is often the rule. I've seen ads and commercials get pulled off
after only one showing or print... that's less than 24 hours.
>Second: if they get enough customers, it really doesn't matter how much they
>piss everyone else off. Consider telemarketers: maybe 95% of people called
>by telemarketers get pissed off, but they don't care, because they get their
>customers from the other 5%.
Tell that to MCI and Sprint. AT&T has had a lot of fun exploiting the
public's response to their competitors calling people constantly to get them
to "switch". Besides, most telemarketers are for relatively small/unknown
products and services... no reputation to lose or parallel publicity
investment to blow if/when they upset people. Personally, I let the
answering machine take care of them... my telephone version of a Kill file...
Now, how this relates to the current discussion (wouldn't want Dick to
cancel this for being "off-charter"): Usenet readers are already well known
for posting the names, phone numbers, and addresses of "people responsible"
for something they don't like. You can bet that if a "Johny Advert" started
up a roboposter to spam newsgroups with junk about a specific product X,
the appropriate names/numbers would appear rather quickly and the people
who hired our pal the poster would hear about it.