Mailing List Archive

On community hostility and the need for change on p5p
As you're probably all aware by now, Sawyer has resigned from the Perl Steering Council (PSC), the Perl Core Team, and other positions he had in the Perl community. He would like to honour his commitment to release 5.34.0, but said he plans to withdraw from the community after that. The reason he gave for his departure was the hostility he has faced in recent years, particularly since announcing the Perl 7 plan at last year's Perl Conference in the Cloud.

Sawyer felt bullied, which is a serious accusation, and one that we should not take lightly. Those who've been accused by Sawyer of bullying claim there was no intention to bully, but that they were engaging in discussion and standing their ground, admitting that they did so strenuously. Furthermore, they believed they had no other choice, because they felt they weren't being listened to, nor their concerns acknowledged.

Setting aside the definition of bullying, the reality is that the style of discourse adopted by people engaging with Sawyer caused him distress and led to him leaving. Sawyer isn’t the only person who has said they’ve stopped participating because of a widespread culture of aggression and derision. This is not acceptable. More though, we're missing out on contributions from people with deep knowledge and experience, and we're undoubtedly putting off new contributors.

Perl, and the ecosystem and community surrounding it, runs almost entirely on the passion of volunteers from around the world. People are passionate because they care about Perl, have invested a lot of themselves in it, and share the same goal: the growth and improvement of the language that they love. But if they become divided because they feel that only their approach will avoid doom, then passions get the better of reason, and by that point the only likely outcome is that it will end in egregious behaviour.

There are wide cultural differences in the way people resolve disagreements. Maybe with friends you have a style of discussion where you get in each other's faces, and the good ideas survive, but that's not viable on p5p, or any forum that wants true representation, and where you're working with people you don't really know. We all need to listen more, and respect the opinions of others. You must be aware of your volume (is anyone else getting a chance to speak?), your tone (are you catastrophizing?), and your manners (are you calling the other person an idiot?).  You should be able to disagree without being rude, and if you can't, we'll call you on it.  And if you persist, you'll be removed.

We cannot afford to keep losing people. Sawyer's departure was very visible and public, but there are others who are quietly walking away, and that's just as damaging, if not more so. Some are (contemplating) leaving because they're burned out, and others are afraid to share their thoughts because of the hostile reception they might get. So we have to change the way p5p works.

Part of the job of the PSC is to pose questions, and then listen to opinions from all quarters, to try and build a consensus, iteratively and transparently. As part of that, we have to ensure that the behaviour of some doesn't suppress the participation of others. Eventually, discussions need to come to an end, and decisions made by the people empowered to do so; this needs to be possible without acrimony.

The definition of Perl 7 is a perfect opportunity for us all to start working on a more constructive and inclusive style of dialogue. We've decided on "use v7", but there's plenty more to work out, and we need to do that together. Please hold fire on your thoughts for this though, until we have shared our thoughts on how to structure this. It would be good to see Perl on the front page of Hacker News for positive reasons, rather than as the poster child for dysfunctional tech communities.

Though we're all volunteers working together on Perl and everything around it, we have to accept that certain standards of conduct are expected of those in leadership positions, whether that's the Perl Steering Council, the Core Team, PAUSE admins, the toolchain, and leaders of other communities within the Perl world. This applies not only to Perl mailing lists, but also on social media, IRC, and elsewhere.

If you have concerns or questions about this, please contact the PSC (steering-council@perl.org) and reflect on this before engaging in discussion publicly. If you do want to reply to this message, please wait at least 24 hours before doing so.

Sawyer became the Perl 5 Pumpking in 2017, and has overseen releases from 5.26 through 5.32 (and hopefully 5.34). This is only part of what he has done for Perl and the community – many of you know him from Dancer and other contributions, and from his energetic and passionate conference presentations. We would like to acknowledge all of his contributions, and thank him for everything that he's done.

We hope that at some point in the future Sawyer, and others, will feel ready and able to return. But there's no chance of that happening until we have made our community a more welcoming and supportive environment. Simply ignoring disagreements and shaking each other warmly by the hand won't work, but neither will shaking each other warmly by the throat. We need to avoid both extremes.

Neil & Rik
Re: On community hostility and the need for change on p5p [ In reply to ]
On 4/16/21 12:52 AM, Neil Bowers wrote:
> As you're probably all aware by now, Sawyer has resigned from the Perl
> Steering Council (PSC), the Perl Core Team, and other positions he had
> in the Perl community. He would like to honour his commitment to release
> 5.34.0, but said he plans to withdraw from the community after that. The
> reason he gave for his departure was the hostility he has faced in
> recent years, particularly since announcing the Perl 7 plan at last
> year's Perl Conference in the Cloud.
>
> Sawyer felt bullied, which is a serious accusation, and one that we
> should not take lightly. Those who've been accused by Sawyer of bullying
> claim there was no intention to bully, but that they were engaging in
> discussion and standing their ground, admitting that they did so
> strenuously. Furthermore, they believed they had no other choice,
> because they felt they weren't being listened to, nor their concerns
> acknowledged.
>
> Setting aside the definition of bullying, the reality is that the style
> of discourse adopted by people engaging with Sawyer caused him distress
> and led to him leaving. Sawyer isn’t the only person who has said
> they’ve stopped participating because of a widespread culture of
> aggression and derision. This is not acceptable. More though, we're
> missing out on contributions from people with deep knowledge and
> experience, and we're undoubtedly putting off new contributors.
>
> Perl, and the ecosystem and community surrounding it, runs almost
> entirely on the passion of volunteers from around the world. People are
> passionate because they care about Perl, have invested a lot of
> themselves in it, and share the same goal: the growth and improvement of
> the language that they love. But if they become divided because they
> feel that only /their/ approach will avoid doom, then passions get the
> better of reason, and by that point the only likely outcome is that it
> will end in egregious behaviour.
>
> There are wide cultural differences in the way people resolve
> disagreements. Maybe with friends you have a style of discussion where
> you get in each other's faces, and the good ideas survive, but that's
> not viable on p5p, or any forum that wants true representation, and
> where you're working with people you don't really know. We all need to
> listen more, and respect the opinions of others. You must be aware of
> your volume (is anyone else getting a chance to speak?), your tone (are
> you catastrophizing?), and your manners (are you calling the other
> person an idiot?).  You should be able to disagree without being rude,
> and if you can't, we'll call you on it.  And if you persist, you'll be
> removed.
>
> We cannot afford to keep losing people. Sawyer's departure was very
> visible and public, but there are others who are quietly walking away,
> and that's just as damaging, if not more so. Some are (contemplating)
> leaving because they're burned out, and others are afraid to share their
> thoughts because of the hostile reception they might get. So we have to
> change the way p5p works.
>
> Part of the job of the PSC is to pose questions, and then listen to
> opinions from all quarters, to try and build a consensus, iteratively
> and transparently. As part of that, we have to ensure that the behaviour
> of some doesn't suppress the participation of others. Eventually,
> discussions need to come to an end, and decisions made by the people
> empowered to do so; this needs to be possible without acrimony.
>
> The definition of Perl 7 is a perfect opportunity for us all to start
> working on a more constructive and inclusive style of dialogue. We've
> decided on "use v7", but there's plenty more to work out, and we need to
> do that together. Please hold fire on your thoughts for this though,
> until we have shared our thoughts on how to structure this. It would be
> good to see Perl on the front page of Hacker News for positive reasons,
> rather than as the poster child for dysfunctional tech communities.
>
> Though we're all volunteers working together on Perl and everything
> around it, we have to accept that certain standards of conduct are
> expected of those in leadership positions, whether that's the Perl
> Steering Council, the Core Team, PAUSE admins, the toolchain, and
> leaders of other communities within the Perl world. This applies not
> only to Perl mailing lists, but also on social media, IRC, and elsewhere.
>
> If you have concerns or questions about this, please contact the PSC
> (steering-council@perl.org) and reflect on this before engaging in
> discussion publicly. If you do want to reply to this message, please
> wait at least 24 hours before doing so.
>
> Sawyer became the Perl 5 Pumpking in 2017, and has overseen releases
> from 5.26 through 5.32 (and hopefully 5.34). This is only part of what
> he has done for Perl and the community – many of you know him from
> Dancer and other contributions, and from his energetic and passionate
> conference presentations. We would like to acknowledge all of his
> contributions, and thank him for everything that he's done.
>
> We hope that at some point in the future Sawyer, and others, will feel
> ready and able to return. But there's no chance of that happening until
> we have made our community a more welcoming and supportive environment.
> Simply ignoring disagreements and shaking each other warmly by the hand
> won't work, but neither will shaking each other warmly by the throat. We
> need to avoid both extremes.

Thank you for this and thank you for how you've been handling it. I am
deeply saddened by it all, but most of all for perl itself.

In addition to this most recent example, I've seen over the years the
flip side; talented individuals who cared deeply, yet due to personal
incompatibilities with others or their own inability to "play nice"
themselves, are no long here through one mechanism or another. I am
saddened by this, as well.

It can be an impossible situation to manage these issues, and sometimes
there is no other choice. Worse yet is the emergence of adversarial
dynamics. Everyone loses in the soul sucking vortex this can produce. It
generates a lot of "collateral damage".

I've thought about this a lot over the years, and I have a general
suggestion that is based fundamentally on 2 things:

1. we all wish to contribute meaningful work, otherwise what are we
doing here?
2. we all love perl/Perl and wish to see it succeed, and in part or in
whole, through our own meaningful contributions

That said, the most successful software projects I have seen have
avoided opportunities for personal conflict altogether (hard to do, no
doubt). Communications happen by way of design documentation, interface
specifications, code reviews, and most of all a well designed
partitioning of tasks.

My suggestion is, just like perl/Perl is meant to be "optimized for
fun"; I think that should be extended to the core development.

Perl development seems to happen very much in a cowboy/bazaar type
style, lines are blurred, technical debates become flash points - or
worse, there is no collaboration at all. The skirmishes break out all
over - email, irc, twitter, git hub, code comments, etc. It's not pretty
when it happens.

We all know, this is unfortunately reality sometimes.

Any development process should accept contributions from people who
themselves may not play well with others or may invoke, for some reason,
the ire of others. The quality and impact of their work should be what
speaks for them and their contributions. We need such "channels" to
exist (and I don't mean just irc or p5p@).

To put it another way; what should matter is that the goal is
accomplished; everyone being best friends or even liking each other,
should not. All that should be required is a way to communicate
technical information formally and having mutual respect (implies also
trust in your peers' abilities). Is this realistic or achievable? I
think so.

How to do it? I don't know. The "perl stack" is far too muddled with
nuance for me to even begin to suggest it. But there are plenty here who
could. And this is probably not all that is needed, but it is needed in
some format.

As always, hope that helps.

Brett

>
> Neil & Rik
Re: On community hostility and the need for change on p5p [ In reply to ]
On Fri, 16 Apr 2021 09:16:53 +0200, B. Estrade <brett@cpanel.net> wrote:
>

I appreciate the energy, but please, let me highlight this request Neil made:

> On 4/16/21 12:52 AM, Neil Bowers wrote:
>> If you do want to reply to this message, please
>> wait at least 24 hours before doing so.

--
With regards,
Christian Walde
Re: On community hostility and the need for change on p5p [ In reply to ]
On 2021-04-16 00:16, B. Estrade wrote:
> Thank you for this and thank you for how you've been handling it. I am
> deeply saddened by it all, but most of all for perl itself.

Agreed. Thank you both Rik and Neil.

> On 4/16/21 12:52 AM, Neil Bowers wrote:
> Sawyer felt bullied, which is a serious accusation, and one that we
> should not take lightly. Those who've been accused by Sawyer of
> bullying claim there was no intention to bully, but that they were
> engaging in discussion and standing their ground, admitting that they
> did so strenuously. Furthermore, they believed they had no other
> choice, because they felt they weren't being listened to, nor their
> concerns acknowledged.

Without excusing Sawyer feeling so bullied that his only escape was to
quit, nevertheless I want to touch on how people "believed they had no
other choice, because they felt they weren't being listened to, nor
their concerns acknowledged". That feeling was so strong that they
raised their voices to "[stand] their ground ... strenuously".

So, my hope is that a take away will be the PSC to look at how more
views can be heard and to ensure those voicing them feel acknowledged.

I further hope that beyond just listening in this email list and perl
dot irc, the PSC will seek to hear what can be done to help the
frameworks and applications through which people use Perl. Be that
BioPerl, Dancer, DBI, Imager, Kahu, Mojolicious, PDL and many others.

Also seeking out the voices of stake holders like businesses and
software distributions like Debian, OpenBSD etc that put perl on to
people's computers.

On 2021-04-16 00:16, B. Estrade wrote:
> Perl development seems to happen very much in a cowboy/bazaar type
> style, lines are blurred, technical debates become flash points - or
> worse, there is no collaboration at all. The skirmishes break out all
> over - email, irc, twitter, git hub, code comments, etc. It's not
> pretty when it happens.

Yes, I feel strongly that the communications channels contribute to the
problem of sometimes casual and obtuse communication.

Email and irc have weak moderation and no means of revising comments
that are misunderstood. It may be prudent to retire them or adjust how
they are run.

For example on this list it may be better to swap settings such that
emails are relayed with senders address removed and only the list
address to reply to. This would discourage unwelcome direct replies off
list.

My preference is irrelevant as I am happy to lurk whilst I am involved
in other aspects of the Perl community.

Nevertheless I feel strongly that the tools are not contributing to
respectful communications while collaborating.


Dean