develooper Front page | perl.perl5.porters | Postings from February 2018

Re: Moderation

Thread Previous | Thread Next
From:
Andreas Koenig
Date:
February 24, 2018 13:14
Subject:
Re: Moderation
Message ID:
87d10upm2n.fsf@k85.linux.bogus
>>>>> On Mon, 12 Feb 2018 20:56:28 +0000, Steffen Mueller <mail@steffen-mueller.net> said:

  > [Resend. There seems to be something horribly wrong in mail
  > post-gsuite-migration email setup and how it interacts with the
  > perl.org email infra. This email was sent on Feb 9th and I'm now
  > giving up on thinking that it's stuck in moderation.]

  > Preface: If anyone following this hasn't seen my previous email in
  > this thread, please read it because it is the actual response from the
  > moderators to Andreas' original question.

  > In the following, I don't speak for any of the other moderators, and
  > I'm describing only my opinions and feelings.

  > Andreas,

  > we've known each other online and from conferences for many, many
  > years now. I have the utmost respect for you, for many reasons.
  > Obviously, your technical and thought-leadership contributions to the
  > community are tremendous. But even more importantly, I've come to know
  > you as a kind and respectful person who is willing to spend a lot of
  > effort on trying to work out what's best for the greater good and best
  > for Perl's users. For example, I very eagerly lapped up your zen on
  > PAUSE take-over requests over a decade ago.

Thank you! And to pass back the compliment, I dearly remember how you
shaped the appearance of PAUSE at that time!

  > It's this context in which it really pains me to say that I feel the
  > attitude that I gleam from your recent emails is rather hostile. Given
  > how (I think) I know you, you must clearly think that there's somebody
  > being wronged* by somebody else's deliberate actions.

Yes, this is true.

  > I feel that you
  > do not trust the set of moderators, and that you do not give us the
  > benefit of doubt about also trying to do the right things.

No, rest assured, there is no reason to believe any of the moderators
has acted in bad faith. I'm talking about symptoms, not people. It's
completely undecided how the symptoms came into being.

  > Considering
  > that we're all here for choice, and in many ways because we're trying
  > to render a free service, as well as the deep respect I have for you,
  > the impression I get from your communication does hurt and
  > disappoint.

The purpose of my mail was to get questions answered that I did not get
answered when I was trying with a normal friendly tone
(https://gist.github.com/anonymous/058d742124c9ab31a308895fcc53aec0). I
was working hard to get rid of my anger, but apparently it leaked
through. I feel very sorry about that.

I do believe we have an unfortunate and dangerous malfunctioning of the
community on multiple levels. It is not easy to find out what needs to
be said and done. And keeping silent is not an option for me, at least
not yet.

Please note that I have chosen to put my doubts into questions because I
saw serious shortcomings that need to be addressed. I dearly thank all
moderators who have committed to step forward and provide their views.
I'll try to clarify my view once again as much as I can. But this will
be my last posting in this thread unless being asked to say something.

  > I believe that the purpose of having moderators - and trust me, each
  > one of us would prefer if none were required - is to ensure that
  > perl5-porters is a place where people want to contribute, feel valued,
  > can have constructively critical but not insulting or hurtful debates,
  > and correspondingly keep coming back, as well as grow as hackers in
  > the process.

I did not question the existence of the moderation group. I repeat
myself when I say I believe it is a good thing to have a working
moderation system. I'd wish we had moderation groups on multiple levels.
Or maybe not. I digress.

  > Now, the line between what's hurtful or not-in-good-faith and what's
  > just curt constructive criticism is very hard to draw. Therefore,
  > moderation needs to be done (sparingly, as you've said yourself in a
  > previous email) by people and cannot be written down as policy and
  > automated by a shell script.

Strongly agreed.

  > I've been a professional manager for a large software engineering
  > department for some time now, and if you'll allow me preach from
  > experience for a brief moment: The single most helpful rule for
  > attaining the above is to always use the most charitable reading of
  > what others say, assume that their intent is good (like yours is, and
  > please believe me, mine), and of course also always maintain good
  > intent oneself. I'd like to ask everyone to take a step back and
  > consider whether they are assuming good intent of their peers,
  > particularly when engaging in meta discussion such as this.

Absolutely. Good vs bad intent is out of question. Like Karl Williamson
suggested, Hanlon's Razor should be the norm. Unfortunately it is
limited in applicability.

  > Thank you for bearing with me. I will try to respond to your questions
  > below. I may run out of time if the kids wake up, and if that is so,
  > I'll send a partial response. I am unlikely to be able to engage in
  > extended further discussion for lack of time and sleep, and I'm doing
  > this in the hope that something is better than nothing. I hope for
  > your understanding.

  > Kind regards,
  > Steffen

  > * There was indeed, Marc, but by error, not evil deliberation.

  > Reminder: This is my opinion only and hasn't been discussed with any
  > of the other moderators.

  > On Fri, Feb 9, 2018 at 6:18 AM, Andreas Koenig
  > <andreas.koenig.7os6VVqR@franz.ak.mind.de> wrote:

 >>>>>> On Wed, 24 Jan 2018 08:14:22 +0100, Andreas Koenig
  >     <andreas.koenig.7os6VVqR@franz.ak.mind.de> said:
    
 >>>>>> On Sun, 21 Jan 2018 20:42:54 +0000, Aaron Crane
  >     <arc@cpan.org> said:
 >>> My fellow list members,
 >>> Please note that, as of commit
 >>> ef7498d2b69936a5c6476ecf1950066f638b2dac, the section of
  >     perlpolicy
 >>> relating to the standards of conduct has been updated.
    
 >> So how did the voting process go? I suppose there was one, and
  >     the
 >> voting moderators are not shy to stand to their decisions,
  >     right?
    
  >     Ping?
    
  >     I have asked some more questions to some members of the community
  >     and
  >     have received the feedback that they are worth being asked. And I
  >     was
  >     told that I should provide some insights about how we could move
  >     forward
  >     from here. I believe the questions themselves partially contain
  >     hints
  >     about a way forward. It all depends on the reaction of the rest of
  >     the
  >     community to determine which ways are reachable.
    
  >     Before I present the questions, I ask the moderators for obvious
  >     reasons
  >     to suspend moderation completely for the duration of this
  >     investigation.
    
    

  > I do not at all find that conclusion obvious.

How much more obvious can it be: we're talking about moderation, the
moderation team forbids questioning the moderation itself, deadlock.
("Any attempts to engage the moderators in further discussion of these
topics, whether on- or off-list, will therefore be treated as actions
warranting moderator intervention." found in
https://www.nntp.perl.org/group/perl.perl5.porters/2018/01/msg248755.html)

Besides: this is a political discussion. Political discussions tend to
turn into ad-personam postings. That was expected and of course it
happened immediately and for that matter I suggested a moratorium. So
far nothing serious has happened that would make it necessary to remind
you of this request, so let's stay it pending.

  > I also feel that your tone is one of an inquisition and find it
  > altogether inappropriate, demeaning. Your use of the word
  > "investigation" implies a sanction that does not exist.

Among the many meanings of the word investigation there are doubtlessly
some that do not have the implication of sanction. We recently had quite
a sour discussion on p5p where English native speakers among themselves
had misconceptions about the term `falsehood` and told each other how to
use a dictionary. I'd prefer not to fall into that trap. Please take the
meaning from the cambridge dictionary, and please pick from the ENUM
"problem", "statement" or "etc." and not "crime":

    the act or process of examining a crime, problem, statement, etc.
    carefully, especially to discover the truth.
    (https://dictionary.cambridge.org/dictionary/english/investigation)

I would indeed like to find the truth. It's the basic requirement that
we all are seeking the truth.

  >     That being out of the way, here they are:
    
  >     1. was the whole group of moderators informed about the state of
  >     the incident with Marc Lehmann's ban?
    
  > Not sure I understand: Do you mean "did all moderators know at the
  > time of Marc's ban that it was to be only for a limited time?" Then
  > yes, we all know as we'd discussed it. Or do you mean "did you all
  > know about how we failed to unban him at the right time?" Then yes,
  > all moderators knew because we discussed that. Without wanting to put
  > words in anyone's mouth, I think I can say that nobody was happy about
  > having missed that (thus also the apology). Note that all of this was
  > before Todd and Karen's involvement.

Thank you!

  >     2. was the whole group of moderators informed about the mail from
  >     former pumpkin Ricardo in which he said that he believes it was an
  >     accident and apologizes and is sure that this will be fixed
  >     quickly?
    

  > Yes, that was after our discussion. None of the moderators can lift
  > the ban ourselves, that's done by the (wonderful and volunteer, I
  > might say) admins. They did so, but I don't think there was an
  > immediate email to p5p (or Marc) about the ban being lifted (typically
  > when multiple people are involved => misalignment on who communicates
  > what).

Thank you!

  >     3. was the whole group of moderators informed why the issue was
  >     not
  >     fixed quickly but was dragged on and on for over three weeks?
    

  > I believe the ban was lifted much quicker than that. I think it took a
  > few days.

Here is the relevant timeline:

2017-12-20

    https://www.nntp.perl.org/group/perl.perl5.porters/2017/12/msg248259.html
    It became public knowledge that Marc was banned. Sawyer knew it,
    nobody else. It was purported to me that Sawyer believed this was a
    legitimate ban. But he says in the posting "for behaviour" as if he
    knew something about the ban. But later when I asked him he did not
    know anything about it.

2017-12-29

    mail from me to Sawyer, meanwhile published as
    https://gist.github.com/anonymous/058d742124c9ab31a308895fcc53aec0,
    asking to set the records straight (not only for Marc Lehmann but
    also for Reini Urban);

    my private and unpublished mail exchange with Sawyer continued until
    2018-01-07; Aaron and Ricardo were put on CC by Sawyer on 2018-01-05;

2018-01-04

    https://www.nntp.perl.org/group/perl.perl5.porters/2018/01/msg248664.html    
    my forwarded mail from Marc Lehmann to p5p

2018-01-12

    https://www.nntp.perl.org/group/perl.perl5.porters/2018/01/msg248755.html
    apology from Aaron Crane to Marc Lehmann; it contained as an
    "explanation" how it had happened: "an administrative oversight
    meant that the ban has been in place since then"

Depending on how you count the days, this is two weeks or three weeks,
but not less.

  >     4. since it was a pumpking who instigated the moderation team, do
  >     the
  >     moderators believe they should report to the pumpking or to
  >     perl5-porters? Or to nobody at all?
    

  > To the pumpking.

  > This being said, if the pumpking (being that Sawyer now or anybody
  > else past or future) said "ye must decide this way or that", I would
  > not blindly follow that order but rather if I felt that it was wrong
  > in a significant way, I would clearly state so, and if need be,
  > resign. In other words, I believe the pumpking delegates the
  > responsibility.

  > I may also add that Sawyer has been extremely responsible about this
  > and has done a stellar job at avoiding any abuse of his influence
  > here. He's stayed out of moderation decisions where he had an
  > involvement in the p5p-incident that preceded the moderation
  > discussion. If he were to get too close to that line, I would be very
  > blunt to him about that, but I know him to be highly sensitive and
  > sensible to this.

Thank you. Frankly, by far I prefer Ricardo's stance who said: "we have
an ethical responsibility to perl5-porters, rather than the pumpking".

(https://www.nntp.perl.org/group/perl.perl5.porters/2018/02/msg249320.html)

  >     5. what was the reason why it took so long to apologize and fix
  >     the ban
  >     settings?
    

  > Order of things as I remember them:

  > a) realization
  > b) head-desking
  > c) lengthy discussion among the moderators about what exactly to do
  > (to avoid any of us accidentally speaking for the others, see also my
  > disclaimers about what I'm doing here)
  > d) apology from Rik

You mean Message-ID: <20180104200601.GA24074@debian>, Date: Thu, 4 Jan
2018 15:06:01 -0500? It was not public. I just happened to be a CC, so I
know about it. But granted, it was an apology, albeit preliminary. And
it was six days after my mail to Sawyer.

  > e) asking admins to unban
  > f) unbanned
  > g) failed or delayed notice about being unbanned as discussed above

  > Now, the times between steps (specifically a to d and d to f) were, I
  > believe a few days each. Do not forget that

I did not forget that. The timeline above still does not look favorable
for the work of the moderation team as a whole.

  >     6. which moderators have access to the system that handles bans?
    
  > None.

  >     7. who has access to the logs of the system?
    
  > I don't believe any of the moderators do.

  >     8. who has investigated, how the ban against Marc Lehmann was
  >     entered
  >     into the system? When and for how long was it entered by whom?
    

  > You know very well who maintains the mailing list services. And it's
  > none of the moderators. Why would you even ask that? I am really
  > trying to give you the benefit of doubt here, but I find this whole
  > thing to be very cynical.

This is really not meant to be cynical at all. I *cannot* ask them
directly. It can only be the moderators' duty to find out what happened
and then to let us know. And most importantly make sure it will not too
easily happen again. I'm a firm believer that I have no right to ask
admins details about the moderation system directly.

  >     9. have the other moderators been informed about the findings?
    

  > "Findings"? What are you even talking about? A ban was put in place
  > way back when. Somebody should have remembered to remove it. (Typical
  > mailing list software - mailman or whatever else - doesn't support
  > automatic-time-dependent-removal of bans as far as I know.) Nobody
  > did.

I'm very grateful that Ricardo investigated more on this matter. Here is
his half finding: "Memory did not serve me right, this ban originated
after the one I had in mind. Nonetheless, if I find out anything, I'll
say so."
(https://www.nntp.perl.org/group/perl.perl5.porters/2018/02/msg249324.html)

And here is Sawyer's finding: "an honest mistake by people who volunteer
for a job that involves literally just the abuse of the community - I
mean, that's really the job description for anyone in the committee."
(https://www.nntp.perl.org/group/perl.perl5.porters/2018/02/msg249315.html)

Read this sentence again and find the hidden messages. It was an "honest
mistake" and the speaker seems to know by whom.

  >     10. which safeguards have been considered and/or implemented to
  >     prevent
  >     that such accidents will not repeat?
    

  > None and I do not think that any are warranted. You do realize that
  > this could have ALSO been fixed by a simple reminder, right?

It *has* started with a simple reminder. But the moderators do not have
a public mail address. So I had to go from one to the other just to find
out that the willingness to unban Marc was close to zero and everything
took far too long. Leave alone the case of Reini who is still waiting to
be unbanned.

Time and again during my private talks I requested that the moderation
team needs to have a common email address. Meanwhile confirmed publicly
by Ricardo (twice), but during my odyssey it was a serious stumbling
block.

When you say, it "could have ALSO been fixed by a simple reminder" I
conclude that nothing of all these mail exchanges was forwarded to you.

  >     11. when did the discussion about the changes to the moderation
  >     rules
  >     that lead to commit v5.27.8-10-gef7498d2b6 take place, who
  >     participated, who voted how?
    

  > Unanimous among the moderators, but before Karen and Todd were added
  > to the group (adding them was also unanimous, I believe).

Thank you.

  >     12. where are the meeting minutes?
    

  > Please. It's an asynchronous email thread among people in different
  > time zones. What kind of shadow organization do you think exists
  > here?

Meeting minutes or a mailbox file, whatever appropriate, show the
arguments brought forth. I mean anything about the process might be
interesting. First of all, what prompted the idea that anything about
the already existing rules in the perlpolicy manpage needed a change *at
all*?

It has been said that the commit message explains the reasoning, but I
find it severly fails to do so. No motive, no impact study.

  >     13. did anybody of the participants in the discussion come to the
  >     idea
  >     that changes to the rules would have to be discussed with and
  >     decided by the people on perl5-porters themselves and not by the
  >     moderators alone?
    

  > I can't speak for the others.

  > I will say, however, that I don't think that this type of discussion
  > makes sense to run on p5p. Mostly the same thing that necessitates
  > even having moderators in the first place. Mailing list membership
  > does not come with some entitlement to vote about anything.

My point is that back room discussions lead to distrust. They should be
kept to a minimum and when the outcome affects the community, they
should be published.

  >     14. how do the moderators think about the idea that they might
  >     have to
  >     obey rules set up by the members of the community, not by
  >     themselves?


  > The moderators are asked to keep things on the list civil by the
  > pumpking. In the past, that duty had fallen on the pumpking alone. Rik
  > then decided to get help and reduce the risk for singular personal
  > bias that way.

  > Do not fall into the trap to equate membership of a mailing list with
  > a sense of community either. It's unfortunately a VERY fuzzy line and
  > we're constantly reminded of that when people refer to "the echo
  > chamber".

I know a bit how difficult it is. I also know that the line is very
fuzzy. My point is not to pretend to know your business better than you.
I just want to remind you (collectively) that you have a responsibility
to the (fuzzy) community and that this responsibility is neglected in
that recent commit.

  >     15. did nobody in the moderation team realize that the new rules
  >     might
  >     be perceived as authoritarian?

  > I cannot speak for the others. But I stated about what I think the
  > purpose of the moderation is. I believe that the current rules are a
  > good tool for that. And even more so, that having people who do their
  > best to make participation in perl development via p5p as constructive
  > and enjoyable as possible. Yes, that requires judgment calls.

I'm questioning the change itself, and I'm questioning how it came
about.

  >     16. did nobody in the moderation team realize that the new rules
  >     might
  >     be perceived as: perl, the superglue of the internet now glueing
  >     the
  >     cylinder lock to lock members out forever?
    

  > Andreas, now you're just being cynical. I feel very disappointed.

What are you now disappointed about? It's eye-catching: glue, lock-out,
cylinder-lock. Perl5-porters has to live with the fact that Reini Urban
is banned and as such victim can tell the world how broken this
community is. And he is always right, at least on this single issue. The
ban was cast for the time of the reign of Ricardo. It should have been
lifted with the new pumpking but was not. I'd say the moderation team is
responsible to fix that, or who else?

My main concern is: how can this be sold as an open community?

And right in that same moment when I write, privately, as a reminder, to
the pumpking that it is time to set the records straight, it is refused
*and* rules are changed simultaneously.

Hanlon's Razor not applicable, at least I need a helping hand that shows
me how.


  >     17. who had the idea that the maximum penalty we have to have on
  >     perl5-porters needs to be indefinite? How long is indefinite
  >     compared to the age of the community? Compared to longest reign of
  >     a
  >     pumpking? Compared to the longest grant paid out by the perl
  >     foundation?
    

  > Do not forget: If a productive, constructive, and civil person chooses
  > to leave perl 5 development because of a toxic environment, that is
  > also indefinite.

Yupp, but it is reversible on their own choice. BTW, the toxic
environment recently stated by several community members did not have
anything to do with verbal abuse neither by Reini Urban nor by Marc
Lehmann. You cannot really call for the moderators when the pumpking
says "CPAN itself is a dumping ground for code. We cannot - and should
not - expect anything written to work ad infinitum"
(https://www.nntp.perl.org/group/perl.perl5.porters/2017/12/msg248288.html)

<satire> Yet it deserves a PAUSE upload and download ban for at least
six lives. </satire>
 
  > So the aim is to balance that as much as possible
  > such that this doesn't happen while bans remain a very rare and dire
  > exception.

I'd venture, in every single case we have seen so far on p5p, a six
month ban would have been sufficient to keep perl5-porters functioning
the way it was (if not better).

  > A single toxic contributor can drive away many others. You
  > can't really win, can you?

Correct. A single toxic of anything can drive away many others. That's
true for commits, moderators, pumkings, discussions. You cannot ban all
of them. The more the back room decisions rule and the less they get
explained, the greater the loss for the community.

  > See also: The trolley problem (except that
  > this isn't congruent in that participants make decisions about their
  > behaviour).

I don't think the trolley problem can be bent to fit the situation.
Everything is reversible, nobody gets killed. Reminds me that it was you
who brought the formula "this is reversible" to all permission take-over
email exchanges on the PAUSE, which was a huge achievement.

  >     18. would you agree that the border between acceptable and
  >     unacceptable
  >     behaviour is not cast in stone and for this reason the moderation
  >     team might decide wrongly, at least occasionally? Which part of
  >     the
  >     moderation system would automatically correct such potentially
  >     wrong
  >     decisions?
    

  > Yes, mistakes can be made. Which is why there's the opportunity to
  > appeal to lift the ban.

The right to appeal is not sufficient for an encounter on par. As I
wrote to Ricardo recently in private mail:

    By opening doors after a limited downtime this would in my humble
    opinion be a valuable standard. Nothing more than that. I would not
    suggest that the reopening of the door should depend on some kind of
    application by the banned person. No "eat humble pie", thanks.

Let alone, that it is subject to random back room diplomacy again, what
happens after somebody appeals for forgiveness. Reminds me too much of
Canossa.

  > There is no automatable and fully
  > deterministic solution to such things and frankly, I think that
  > anybody who would seriously argue that this can be automated (which
  > you haven't done, I might add, but you seem to imply it by the
  > phrasing of the question) clearly can't be serious or somehow has lost
  > touch with other human beings.

My god, no, I was not talking about scripting. I was talking about the
fact that the previous rules were better. A maximum ban of six months
and reset of the warnings state after that time was *the way* to
automatically correct nearly every mistake by any moderation team,
whatsoever. There was a certain amount of wisdom in that previous text.

Having a moderate maximum ban length forces the community to face the
critics from outside the echo chamber within their own premises
occasionally. That's a good and healthy opportunity for everybody within
the echo chamber to think about their own responsibilities. A good
pumpking should have the wider community in his mind and a good
moderation team should stay inviting on the occasional need to issue
warnings or bans.

A good moderation team should never speak like a supreme court (but all
recent postings had some such hostility), should always explain why they
think what they are saying is correct (all postings either did not or
did only after being questioned) and should remind that doors are in
principle open (missing in all postings).

When you see a sentence like "The wording of the forwarded message is
clearly in contravention of our standards of conduct", you should expect
some substantiation. When it is missing, something is very wrong. And it
was missing.

Especially the idea to extend moderators' power to unlimited off-list
world is extremely daunting.

You see what a minefield (term coined by Marc Lehmann) you are setting
up? From freely speaking about the weirdness of the policy change, now
all of a sudden I am possibly guilty "to have a moderator engaged in
further discussion" and as I turn around, you have the right to ban me.
Even if I post this somewhere else.

When I could choose between a p5p without moderation and this minefield,
I knew where to go.

  >     While it may sound ungrateful when I present such a shitload of
  >     questions, I'd like to let you know that I'd rather not have had
  >     to go
  >     into this endavour. But it's not about my personal inclination
  >     when I
  >     set up this catalogue. I hear the questions and I have to spell
  >     them out
  >     because apparently nobody else does, while still they actually are
  >     awaiting to be dealt with.
    
    

  > I do find your questions to be overbearing, seemingly ungrateful, and
  > phrased in a very unfortunate way, likely out of great frustration. I
  > feel the questions are accusatory, do not give any benefit of doubt,
  > show no assumptions of good intentions, and appear to show many
  > underlying assumptions on your part that I would not agree with if
  > spelled out.

I hope I could relativize some of your perceptions. I'm sorry that I did
not find a less accusatory wording to let you all know how I perceive
the recent development. My biggest mistake by far was to engage in
private discourse about the thing. I can only recommend to everybody
ever coming into this sort of personal/political challenge, go public as
quickly as possible, or better yet start it publicly.

But do it before you get banned. Once you are banned you need a proxy
and those are in short supply at the moment. So speak up now. You still
can prevent your own future demise.

  > The style has made me want to sever my ties to perl
  > development altogether since I'm sure to feel miserable about this for
  > some time to come. I'm quite convinced I'm not alone in this dread and
  > some others would be a much bigger loss than I by far. Being involved
  > in p5p is a hobby for me. As such, it needs provide something positive
  > to warrant the time and energy spent. Maybe a feeling of community and
  > friendship. Maybe a sense of achievement. Maybe a sense of comfort
  > about social interaction. You name it. Whatever it is if the
  > involvement makes me feel dreadful and drained, that has an impact on
  > more important thing such as my ability to give my best to my family,
  > or to have maximum empathy available for looking after others whose
  > career I influence in my work.
  > Why do I expand on what this means for me so much? Because the way I
  > feel is exactly what I believe should be avoided inflicting others
  > because it's sure to drain our ability to keep perl5 healthy.

Amen.

    But the possibility of abuse may be a good reason for leaving
    capabilities out of other computer languages, it's not a good reason for
    leaving capabilities out of Perl.
                 -- Larry Wall in <199709251614.JAA15718@wall.org>

  > I hope my (personal) responses above are helpful.

You (and all the others who contributed creatively and constructively to
this thread, publicly or privately [you know who you are!]) helped me a
lot to revitalize a tiny inner believe that not all is lost. Thanks a
lot for that.

  > Kind regards,
  > Steffen

  >     Take care,
    
Thanks again,
-- 
andreas

Thread Previous | Thread Next


nntp.perl.org: Perl Programming lists via nntp and http.
Comments to Ask Bjørn Hansen at ask@perl.org | Group listing | About