Article <vr2bqu$o3d$1@reader1.panix.com>

Deutsch English Français Italiano
<vr2bqu$o3d$1@reader1.panix.com>

View for Bookmarking (what is this?)
Look up another Usenet article
Path: ...!weretis.net!feeder9.news.weretis.net!panix!.POSTED.panix6.panix.com!nan.users.panix.com!robomod!not-for-mail
From: pschleck@panix.com (Paul W. Schleck)
Newsgroups: news.groups.proposals,news.groups
Subject: Re: 3rd RfD: Mass-deletion of moderated groups without a moderator
Date: Fri, 14 Mar 2025 19:58:41 EDT
Organization: PANIX Public Access Internet and UNIX, NYC
Approved: NGP Approval Key <ngp-approval-key@ngp.big-8.org>
Message-ID: <vr2bqu$o3d$1@reader1.panix.com>
References: <vqq7tf$ehn$1@reader1.panix.com> <vqtir0$2ukis$1@dont-email.me>  <vqujqc$bid$1@reader1.panix.com> <vquse4$3d1t6$1@dont-email.me>  <vr1b28$at4$1@reader1.panix.com> <vr1src$1rj2d$1@dont-email.me>
Injection-Info: reader1.panix.com; posting-host="panix6.panix.com:166.84.1.6";
	logging-data="22083"; mail-complaints-to="abuse@panix.com"
User-Agent: nn/6.7.3
Bytes: 5045
Lines: 61

In <vr1src$1rj2d$1@dont-email.me> D Finnigan <dog_cow@macgui.com> writes:

>On 3/14/25 10:16 AM, Paul W. Schleck wrote:
>>
>> 
>> Such a service would realistically have to employ monitoring/alerting,
>> SPAM filtering, keyword trapping, duplicate detection, rate limiting,
>> and manual review of any queued articles for false positives. This is
>> starting to resemble the duties of a human moderator, and a significant
>> workload for one volunteer long-term.
>>

>Are you volunteering to run the robo-moderator?

No, Rayner did, and he asked for advice:

In <vqtir0$2ukis$1@dont-email.me> Rayner Lucas <usenet2025@magic-cookie.co.ukNOSPAMPLEASE> writes:

>For groups in the latter category, I'm considering setting up some kind of 
>robo-moderation service for them. This would have a couple of benefits: 
>it would give time to try converting a group to unmoderated as a test 
>case, and would also permit seeing whether anyone is still attempting 
>to post to the groups. It could therefore serve as a temporary measure 
>if it's unclear what the best course of action would be.

>A robo-moderation system could also be a starting point for a more 
>general moderation platform. Currently, a serious problem is that 
>prospective moderators can't simply start moderating a group: they need 
>to set up email addresses, install and configure software (most of which 
>is outdated and awkward to set up), and get their Usenet provider to 
>allow them to post approved messages (which not all providers will be 
>willing to do). If we're going to have a mass deletion of groups without 
>moderators, I think we also ought to make sure that moderating a group
>is not an unreasonably difficult thing to start doing.

>Thoughts?

>R

Personally, I think this robo-moderation idea has the risk of being a
lot of effort for little reward.  However, in the spirit of constructive
feedback, since it was solicited, I tried to walk everyone through some
use-case/role-playing exercises to see how this might work out in actual
practice.  I wanted Rayner and the rest of the group to consider if they
really wanted to go down this road, and if they realistically have the
time, interest, and resources to make it succeed.  I also wanted
everyone to be mindful of past failures and consider how this latest
effort would mitigate them.

For example, the Panix STUMP infrastructure that the Big-8 Board uses to
moderate its newsgroups (including news.announce.newgroups) could be
pressed into service for the robo-moderation gateway.  It can already do
monitoring/alerting, SPAM filtering, keyword trapping, duplicate
detection, and manual review of any queued articles for false positives.
It currently does not do rate limiting, but that could be easily added
as a new automatic rejection category with configuration settings and
some scripts.

--
Paul W. Schleck
pschleck@panix.com