Requesting a cool down period.

Requesting a cool down period.

[Carlos O'Donell]

As a GNU project maintainer for the GNU C Library, I would like
to request a cool down period in discussions relating to the
abort() warning cartouche, the commit to remove it, and the
reversal. That is to say that within the next 24 hours we stop
writing and responding in order to cool down.

There are only 7 weeks until the ABI freeze for glibc 2.28,
only 11 weeks until the release, and we owe our users the best
release possible.

I suggest that the cool down period on this topic last until 
August 1st, *after* we release glibc 2.28. Yes, this means the
2.28 release will include the text in the cartouche, but in this
community we don't need to rush things, and we follow consensus.

I know that this request may disappoint some of you. We should
not remove the abort() warning cartouche until we have consensus
again, and it should be well summarized. This community abides by
consensus rules, and consensus on this issue may take long enough
to tally that it may disrupt the release of 2.28, and for the sake
of our users we should not allow that to happen.

I suggest we all take the next ~11 weeks to cool down, reflect,
study, consult with others, and come back with ideas related to
both the abort() warning cartouche and project governance.

Please lend me your support by abiding by a cooling down period,
and preparing for a proper discussion after the cool down.

Please also lend me your support in making 2.28 the best glibc
release possible.

-- 
Cheers,
Carlos.

Re: Requesting a cool down period.

[Siddhesh Poyarekar]

On 05/10/2018 11:43 AM, Carlos O'Donell wrote:
> Please lend me your support by abiding by a cooling down period,
> and preparing for a proper discussion after the cool down.

+1, does having a mojito for breakfast count as having already cooled 
down or should I have another one? ;)

Siddhesh

Re: Requesting a cool down period.

[Federico Leva (Nemo)]

(Replying because I was in cc.)

Carlos O'Donell, 10/05/2018 09:13:
> As a GNU project maintainer for the GNU C Library, I would like
> to request a cool down period in discussions relating to the
> abort() warning cartouche, the commit to remove it, and the
> reversal. That is to say that within the next 24 hours we stop
> writing and responding in order to cool down.

That's sensible. Personally I'm already applying the ReactLater pattern 
(<https://web.archive.org/web/20170325222358/meatballwiki.org/wiki/ReactLater>) 
and I can easily keep my responses in draft for an even longer time.

Federico

Re: Requesting a cool down period.

[Mark Wielaard]

On Thu, May 10, 2018 at 02:13:07AM -0400, Carlos O'Donell wrote:
> I suggest we all take the next ~11 weeks to cool down, reflect,
> study, consult with others, and come back with ideas related to
> both the abort() warning cartouche and project governance.
> 
> Please lend me your support by abiding by a cooling down period,
> and preparing for a proper discussion after the cool down.

I think this is sensible given how heated the discussion still is
and now involves separate issues that all could use a reset.

Thanks,

Mark

Re: Requesting a cool down period.

[Torvald Riegel]

On Thu, 2018-05-10 at 02:13 -0400, Carlos O'Donell wrote:
> Please lend me your support by abiding by a cooling down period,
> and preparing for a proper discussion after the cool down.

I think we should still point out behavior on the mailing list that we
think is not representative of the glibc community, and not accepted by
the community (eg, the anonymous comment I replied to).  We might have
many new people reading the list or just the particular thread in
question, who haven't sampled all other posts on the list.  I'd like to
avoid that they get wrong impressions.

Re: Requesting a cool down period.

[Carlos O'Donell]

On 05/10/2018 08:22 AM, Torvald Riegel wrote:
> On Thu, 2018-05-10 at 02:13 -0400, Carlos O'Donell wrote:
>> Please lend me your support by abiding by a cooling down period,
>> and preparing for a proper discussion after the cool down.
> 
> I think we should still point out behavior on the mailing list that we
> think is not representative of the glibc community, and not accepted by
> the community (eg, the anonymous comment I replied to).  We might have
> many new people reading the list or just the particular thread in
> question, who haven't sampled all other posts on the list.  I'd like to
> avoid that they get wrong impressions.
 
I agree. This would not be a discussion of the above topics, but hopefully
a brief message that the topic is in cool down, and that the poster's
behaviour is is not acceptable. Some people, like strangers, will not abide
by the cool down, and should be reminded gently.

-- 
Cheers,
Carlos.

Re: Requesting a cool down period.

[Alexandre Oliva]

On May 10, 2018, "Carlos O'Donell" <carlos@redhat.com> wrote:

> On 05/10/2018 08:22 AM, Torvald Riegel wrote:
>> On Thu, 2018-05-10 at 02:13 -0400, Carlos O'Donell wrote:
>>> Please lend me your support by abiding by a cooling down period,
>>> and preparing for a proper discussion after the cool down.
>> 
>> I think we should still point out behavior on the mailing list that we
>> think is not representative of the glibc community, and not accepted by
>> the community (eg, the anonymous comment I replied to).  We might have
>> many new people reading the list or just the particular thread in
>> question, who haven't sampled all other posts on the list.  I'd like to
>> avoid that they get wrong impressions.
 
> I agree. This would not be a discussion of the above topics, but hopefully
> a brief message that the topic is in cool down, and that the poster's
> behaviour is is not acceptable. Some people, like strangers, will not abide
> by the cool down, and should be reminded gently.

May I suggest (excuse the one-past-the-end post) that, at least during
the cool-down period, we refrain from even pointing out what is or isn't
acceptable tone?

One reason is that this debate is so heated that it's far too easy to
mistake dissent for inadequacy of tone.  Indeed, AFAICT in 100% of the
cases in which one party criticized the tone of the other party, the
parties held opposite positions in the debate.  What's more: the same
words, when used against or for people holding a certain position, got
praise or criticism, respectively, depending on whether the
praiser/critic's opinion was aligned or misaligned with the target of
the praise or criticism.  We don't want to even look like we're
suppressing dissent, especially for strangers.


Another reason is that opinionated criticism (as in, criticism that
carries with it an opinion on the subject matter) is provocative to
dissenters, so it achieves the opposite of the intent of the cool-down.


One way to deal with these issues, especially the former, would be to
require that whoever criticises an opinion on its tone MUST hold the
same position as that expressed in the criticized opinion.  That would
remove the appearance of opinionated or biased criticism within regular
participants.  However, the very strangers that might be targets of such
criticism might very well not be aware of the critic's position, so they
might still honestly feel that the criticism is a disguised attempt at
suppression of dissenting opinions.


So it seems to me that, if we're to post such messages, it would be best
if they were limited to a link to the cool-down request, in private, and
nothing else whatsoever.


Thanks for your consideration,

-- 
Alexandre Oliva, freedom fighter    http://FSFLA.org/~lxoliva/
You must be the change you wish to see in the world. -- Gandhi
Be Free! -- http://FSFLA.org/   FSF Latin America board member
Free Software Evangelist|Red Hat Brasil GNU Toolchain Engineer

Re: Requesting a cool down period.

[Carlos O'Donell]

On 05/10/2018 01:29 PM, Alexandre Oliva wrote:
> So it seems to me that, if we're to post such messages, it would be best
> if they were limited to a link to the cool-down request, in private, and
> nothing else whatsoever.

The requests should be made in public. I want to avoid the appearance
that the glibc community allows toxic posts by random developers to go
without warning. A private response could make it appear like we don't
care, or worse endorse the view.

Feel free to end any of your current discussions with a link to the
cool down request. I did not want to make the cool down request overly
broad.

-- 
Cheers,
Carlos.

Re: Requesting a cool down period.

[Carlos O'Donell]

On 05/10/2018 02:01 PM, Carlos O'Donell wrote:
> On 05/10/2018 01:29 PM, Alexandre Oliva wrote:
>> So it seems to me that, if we're to post such messages, it would be best
>> if they were limited to a link to the cool-down request, in private, and
>> nothing else whatsoever.
> 
> The requests should be made in public. I want to avoid the appearance
> that the glibc community allows toxic posts by random developers to go
> without warning. A private response could make it appear like we don't
> care, or worse endorse the view.
> 
> Feel free to end any of your current discussions with a link to the
> cool down request. I did not want to make the cool down request overly
> broad.
 
... when I say "made in public" I mean specifically:

CC'ing libc-alpha@sourceware.org.

-- 
Cheers,
Carlos.

Re: Requesting a cool down period.

[Torvald Riegel]

On Thu, 2018-05-10 at 14:29 -0300, Alexandre Oliva wrote:
> On May 10, 2018, "Carlos O'Donell" <carlos@redhat.com> wrote:
> 
> > On 05/10/2018 08:22 AM, Torvald Riegel wrote:
> >> On Thu, 2018-05-10 at 02:13 -0400, Carlos O'Donell wrote:
> >>> Please lend me your support by abiding by a cooling down period,
> >>> and preparing for a proper discussion after the cool down.
> >> 
> >> I think we should still point out behavior on the mailing list that we
> >> think is not representative of the glibc community, and not accepted by
> >> the community (eg, the anonymous comment I replied to).  We might have
> >> many new people reading the list or just the particular thread in
> >> question, who haven't sampled all other posts on the list.  I'd like to
> >> avoid that they get wrong impressions.
>  
> > I agree. This would not be a discussion of the above topics, but hopefully
> > a brief message that the topic is in cool down, and that the poster's
> > behaviour is is not acceptable. Some people, like strangers, will not abide
> > by the cool down, and should be reminded gently.
> 
> May I suggest (excuse the one-past-the-end post) that, at least during
> the cool-down period, we refrain from even pointing out what is or isn't
> acceptable tone?

I don't think this is needed.  For all posters adhering to the
cool-down, there's no issues because there are no posts by them.
Then there are posts such as the two coming from the guerillamail.com
accounts, which need to be responded to, IMO.

> One reason is that this debate is so heated that it's far too easy to
> mistake dissent for inadequacy of tone.  Indeed, AFAICT in 100% of the
> cases in which one party criticized the tone of the other party, the
> parties held opposite positions in the debate.  What's more: the same
> words, when used against or for people holding a certain position, got
> praise or criticism, respectively, depending on whether the
> praiser/critic's opinion was aligned or misaligned with the target of
> the praise or criticism.  We don't want to even look like we're
> suppressing dissent, especially for strangers.

I disagree with your description of the situation, but now we're getting
back into the topics we didn't want to discuss, so I won't comment in
more detail...

Re: Requesting a cool down period.

[Waldemar Brodkorb]

Hi Carlos, Hi all,
Carlos O'Donell wrote,

> As a GNU project maintainer for the GNU C Library, I would like
> to request a cool down period in discussions relating to the
> abort() warning cartouche, the commit to remove it, and the
> reversal. That is to say that within the next 24 hours we stop
> writing and responding in order to cool down.
> 
> There are only 7 weeks until the ABI freeze for glibc 2.28,
> only 11 weeks until the release, and we owe our users the best
> release possible.
> 
> I suggest that the cool down period on this topic last until 
> August 1st, *after* we release glibc 2.28. Yes, this means the
> 2.28 release will include the text in the cartouche, but in this
> community we don't need to rush things, and we follow consensus.
> 
> I know that this request may disappoint some of you. We should
> not remove the abort() warning cartouche until we have consensus
> again, and it should be well summarized. This community abides by
> consensus rules, and consensus on this issue may take long enough
> to tally that it may disrupt the release of 2.28, and for the sake
> of our users we should not allow that to happen.
> 
> I suggest we all take the next ~11 weeks to cool down, reflect,
> study, consult with others, and come back with ideas related to
> both the abort() warning cartouche and project governance.
> 
> Please lend me your support by abiding by a cooling down period,
> and preparing for a proper discussion after the cool down.
> 
> Please also lend me your support in making 2.28 the best glibc
> release possible.

That is a really good idea and I sent over 3 euro/dollar to every
libc developer/maintainer (including RMS and Rich Felker, just to be
fair) involved in this discussion via Paypal so that you get an ice
cream, chill in the sun and cool down!

Sunshine,

 Waldemar, just a user and regular consumer of GNU C Library code...

P.s.: just sent me a personal mail to wbx@openadk.org if you want to
get your ice cream.

Re: Requesting a cool down period.

[Javiera Serrano Polo]

[-- Attachment #1: Type: text/plain, Size: 438 bytes --]

El dj 10 de 05 de 2018 a les 02:13 -0400, Carlos O'Donell va escriure:
> As a GNU project maintainer for the GNU C Library, I would like
> to request a cool down period in discussions relating to the
> abort() warning cartouche,

Since there is work in progress regarding this matter and targeted
before release, I object.

> Please also lend me your support in making 2.28 the best glibc
> release possible.

This is the idea.

[-- Attachment #2: smime.p7s --]
[-- Type: application/pkcs7-signature, Size: 3386 bytes --]