Continuing my thoughts about listadmin best practices.  We've covered:

o  Listadmin sanctions should be carefully minimal.  (Often as not, the
   misbehaving person is seeking public drama and wants nothing quite
   a much as to be treated harshly, towards a goal of claiming 
   martyrdom at the end.)

   There's a saying that applies:  'Do not feed the energy beast.'

   I will comment about the range of available sanctions later.

o  Listadmin sanctions & related discussion should ideally be on the
   record, somewhere public but not with an inherent huge audience
   (like a main mailing list).  Carrying out intervention against
   problem behaviour on the main forum creates a (sometimes bigger) 
   mess, and (again) see prior point about drama-seeking and 
   martyrdom-seeking.

   Above is the single point I hear denied most strenously by my 
   fellow listadmins, under a variety of excuses and reasons, such 
   as 'We've found it better to conduct these matters privately'
   (leaving in question the key term 'better'), and 'We must act
   only privately to protect people's privacy (which is nonsense
   in this context), or 'Criticism should always be in private'
   (which begs the question of whether transparency doesn't have 
   significant value).  What this really amounts to, basically, is
   the person thinking 'I'm used to doing it this other way, and
   I'm not going to change just because you have an opinion.'

   The speaker typically doesn't wish to think about downsides of 
   the out-of-band-in-private-mail practice, including the 
   _perception_ (and sometimes the reality) of being another Pete Salzman.

o  Listadmin sanctions should be immediately disclosed to the erring
   member, cc'd to the public record, _and_ should have a preannounced
   duration, _and_ expiration of that time period should again be
   politely disclosed to the person CC'd to the public record.

   A small but highly public mailing list operated for administrative
   purposes turns out to make a pretty good public record for these
   purposes, especially if you stress to the misbehaving person that 
   he/she is entirely free to discuss the matter there and _not_ 
   subjected to any sanctions for so doing.

   The existence of such a public record means your -actual- 
   de-facto enforcement can be confirmed and observed without 
   suspicions of convenient inconsistencies and undisclosed 
   personal vendettas.

o  Be as nice and have as relaxed/indulgent an attitude as you can
   manage, both as a goal in itself and to help avert escalation.
   (People being criticised tend to double down out of anger.)

o  As many hard rules as possible should be automated and applied by 
   the mailing list software.  This dictum emerged from something
   I noticed and was surprised by:  If as listadmin you ask members,
   however politely, to cease doing something obviously wrongful, 
   they give you endless backtalk and bad attitude, but if you 
   configure GNU Mailman to make said misbehaviour unsuccessful, 
   they perceive this as 'just the way the mailing list works' 
   rather than a rule applied by humans and somehow open to debate
   with them getting a vote.

   For example, I mentioned the half-dozen main trolls on the main
   SVLUG mailing list having a habit of escalating any discussion
   of listadmin sanctions by cross-posting their backtalk to additional 
   mailing lists.  I tried the gentle 'Please don't do this' approach
   exactly once before adding Mailman spam filters causing any future
   crossposts matching that pattern to be autorejected.

   But before that, for quite a few months, several posters kept 
   cross-posting for other reasons, most notably a past SVLUG VP, 
   Mr. Bill Ward.  I politely asked Ward (and the others) to 
   avoid those cross-posts because they caused considerable extra
   manual work in the held-messages queues for each of the mailing 
   lists' listadmins as soon as subscribers of one or the other
   lists attempted cross-posted replies without being subscribed
   to all of them.  Ward (most particularly) angrily asserted his
   right to keep misbehaving, and felt that he need not respect 
   my and the other listadmins' time (which, after all, we were
   giving away for free, so how dare we object to it being wasted).

   I didn't further argue with Ward, because I _did_ value my time.  ;->
   However, I noticed that after I created the anticrossposting filters
   to foil the trolls' escalation game, not only did none of them 
   try to complain and argue against those, but neither did Ward and
   the other habitual crossposters.

   My best guess:  Their not seeing a human to argue with resulted
   in them classifying this change as 'just how the mailing list works' 
   rather than as a 'rule', where they felt a moral entitlement to 
   bicker with and 'vote' against the latter but not the former.

   A quick look through GNU Mailman's administrative screens will show a
   listadmin a large number of rules implemented as
   just-how-the-software-works, e.g., total message size limit, which
   defaults to 40kB., total permissible number of To: and Cc: addressees
   without landing in the held-messages queue (defaults to 10, but 
   appears to kick in a bit under that number), and many more.

   Is it cynical and manipulative to embed into software your 
   mailing list rules in the knowledge that people won't argue with 
   them?  Maybe so.  But it sure works.  ;->


Something I used to say, but this turns out to be a subtle matter:

o  Apply only the published rules, without bias, and no others.
   Do not punish someone for violating a rule you just pulled out 
   of /dev/ass.

The above was my first and most important principle as listadmin, when 
SVLUG mailing list administration was suddenly handed to me around 2007.
I publicly vowed that the documented rules and _only_ the documented
rules would be applied, without bias, against all participants starting
with myself.

There was a reason for this:  SVLUG had a bad history in this area, 
e.g., my friend Marc Merlin, who was then President and listadmin, 
said on the main mailist list in 2001 that henceforth any subscriber
who stated that another subscriber's posting had been incorrect in any
way, in no matter how cordial a manner, would be permanently banned
without warning.  I asked if he were quite serious, and he confirmed
that he was.  I and several other prominent members responded by quietly
ceasing to participate and take our contributions elsewhere, mostly to 
_Linux Journal_ editor Don Marti's linux-elitist mailing list and
CABAL's discussion list.  

Several times, I asked President Marc Merlin and his successor President
Don Marti whether the alleged permaban rule was still in force (yes), 
and then also whether they were going to document this rule on the rules
Web page, http://www.svlug.org/policies/list-policy.php (no).  I noted 
this ongoing cheeky hypocrisy on the BALE calendar page's entry for
SVLUG (see:
http://web.archive.org/web/20040624133857/http://linuxmafia.com:80/bale/#svlug)
, but stayed away from SVLUG's mailing list until Don Marti's term
of office expired in 2004.

Many years later, I learned that an extremely passive-aggressive SVLUG
member who worked at Cisco Systems had threatened cancellation of
SVLUG's then monthly meeting space at Cisco unless SVLUG's officers 
curtailed anyone on the main discussion list being permitted to 
contradict others' factual assertions, as he felt that the excessive
focus on technical accuracy was 'driving away novice users'.

This was all in the background (except for the revelation about the
gentleman from Cisco) when I inherited mailing list administraiton, and
vowed to make a clean break from backroom politics and from arbitrary and
incomprehensibly harsh sanctioning of people who'd broken no documented
rule.  So, I did that.

The first thing that happened was that I wasn't doing anything about 
the half-dozen trolls repeatedly overwhelming the main mailing list with
dumb arguments and attempts to provoke reactions.  Some regulars wrote
to me, complaining that I wasn't taking action.  I replied, would you
have me be a hypocrite?  These people are being annoying, but there's no
rule they're transgressing.  Why weren't the complainers just killfiling
what they didn't want to see?, I asked.  But I said, give me a few hours
to think this over.

I thought hard about the wording I said was my policy's ethos:

   SVLUGs' listadmins normally intervene only [snip spam bit]
   and to halt major eruptions of offtopic spew.

By happy accident, the latter phrase was something I could work with.
Although at the time I wrote that phrase, what I had in mind was someone 
making (e.g.) a poetrybot post nonsense text to the mailing list, the 
wording also naturally implied other things, too -- like 'offtopic spew'
that was of sufficient volume, duration, and obnoxiousness to impair the
mailing list as a whole.

I expounded on this slightly new interpretation of policy:  The
listadmins would not be content police, _but_ would intervene to 
curb any offtopic traffic significant in size, duration, noxiousness,
etc. to impair the mailing list as a whole.  The listadmins' job isn't
to make _you_ (any individual subscriber) happy.  If you
ididosyncratically hate some person or some topic, just killfile it.
The listadmins were not going to sanction some other guy for your
personal benefit.  The most we'd do in that case was give you some free
killfile-administration lessons.  The listsadmins' job is to keep the
mailing list as a _whole_ healthy.  They would deal only with
systemic-level injuries.

At this point, I happily advised the half-dozen combatants that they had
better simmer down or would find themselves gently and if necessary
less-gently sanctioned.  They tried to provoke overreaction:  This
didn't get a rise from me.  They tried to escalate to massive
cross-posting.  I ended that with filtering rules.  And then, it not being
fun any more, and no public turmoil or angry reaction or flamboyant
martyrdom being on offer, they gave up and went elsewhere.  Mirabile
dictu!

Some amateur-lawyer types eventually thought to object that some of
this phrasing was suspiciously vague, e.g., that nothing declared in
advance what would and would not be deemed 'spew', and that the term
'offtopic' was left undefined.  Again, this was a happy accident, but,
by the time they thought to make this complaint, I had pondered the
matter and had my answer:

Yes, it's just a little vague.  Clever of you to notice that.  No, we're
not going to fix that.  No, we're absolutely not going to paint a
figurative yellow line down the border between spew and non-spew, for 
lots of reasons including not giving a bunch of OCD edge-case maniacs 
and Asperger's Syndrome poster-children  a target to shoot at.   

Yes, 'offtopic' is not specifically defined, nor 'major eruptions'.
Care to guess why?  It's because we figure competent adults can and
should figure that out.  If you cannot, you might end up having
listadmin encounters, but shouldn't worry too much about that, because,
as it says on the policies page,

   Enforcement if any should always be minimal and public. (We don't do
   backroom politics, and our preferred means of social control is to
   help everyone apply his/her own well-tuned killfile.)

So, if you're unsure what's offtopic, don't sweat it too much.  For one
thing, every mailing list including ours deliberately indulges a
moderate amount of offtopic banter.  Listadmins aren't going to react 
unless / until its prominence / volume / obnoxiousness rises to
impairing-forum-as-a-whole 'major eruptions of offtopic spew' level.
So, don't sweat it.  You're probably fine.  If not, we'll all learn, and
it will be a gentle promise.  If we're lying, you're free to call us
hypocrites, but be aware that we're putting everything on the public
record so you embellish the truth at some peril of being proven wrong
and mocked for it.

And this has been what we've done ever since 2007, and it's worked
great.  And it turned out, on reflection, that _no_ mailing list 
ever refuses to take sanctions unless the listadmin can point to 
violation of a specific rule.  That's actually something of a children's
fantasy idea of governance.  Consider:

SVLUG's mailing list doesn't have a rule saying 'No spam allowed.'  It 
doesn't have one saying 'No Ponzi scheme criminal marketing allowed.'
Why?  Because it turns out that some things just go without saying.

No sane listadmin has the attitude 'We can't sanction you unless we hae
a specific rule against what you're doing', because that's absurd.
There's no way you could list every possible type of misbehaviour, nor
should you.

Prior to my taking over SVLUG listadmin and sysadmin duties, the
officers _were_ dumb enough to try.  Have a look at the policies page
before I took an axe to its bloatedness:
http://web.archive.org/web/20040202102133/http://www.svlug.org:80/policies/list-policy.shtml
Have you ever seen anything quite so boneheaded?  We don't do that, any
more.



o   Know your all options for listadmin sanctioning, and choose whatever
    is the least drastic one available to make your point.

    I've mentioned several:
    o  Mild reproval (taking care to speak explicitly in listadmin capacity)..
    o  Moderated flag for a preannounced limited period.
    o  Moderated flag _and_ no posts to be approved for a preannounced limited 
period.
    o  Either of the prior two until the person posts a public apology.

    Prior SVLUG listadmins, but not me, used:
    o  Threat of permanent banning instantly upon future offence.

    Variations:
    o  Removal without ban.
    o  Removal and ban (permanent or not).

    On a different mailing list, I've used 'removal without ban' twice,
    and I'll clarify.

    In both cases, the subscriber indulged quite severe misbehaviour
    that he knew would be particularly objectionable.  For policy
    reasons, I wanted to both be gentle _and_ to get the person's
    attention.  So, I wrote the subscriber, CC to the mailing list,
    saying I had just removed that person's subcription but did not
    otherwise sanction him and would in no way prevent him from
    resubscribing.  I was making an explicitly dramatic statement,   
    saying 'Hey, that was very not OK, and I could easily have removed 
    you and locked you out, but I don't want to do that without clear
    need, so I'm getting your attention by making it necessary for you
    to rejoin before you can post again.'

    People tend to get the message that you are making a strong
    statement but are being relatively kind, when you do this.

    Again, putting the action on the record strengthens the message 
    by making clear that there _is_ enforcement and that it occurs
    even-handedly without any arbitrary aspects.


I really do strongly recommend all of the above guidelines.

_______________________________________________
Dng mailing list
Dng@lists.dyne.org
https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng

Reply via email to