[llvm-dev] RFC: Introducing an LLVM Community Code of Conduct

Lang Hames via llvm-dev llvm-dev at lists.llvm.org
Wed Oct 14 15:24:36 PDT 2015


 >> (For the record, I'd normally not have said anything, but since this is
specifically in a thread about community social norms...)
> You should say something, even normally. =]

Well alright then... ;)

> However, it is incredibly important to not expect or demand that a person
*you have made feel unsafe* take the time to explain why. Being forced to
explain the problem can and in many cases does *exacerbate* the problem for
the individual. If they are up for the challenge of explaining, wonderful.
But if they are not, it is critical to respect that and give them the space
to feel safe again. Talk to your friends, to others, to the advisory
committee, to whomever else you need to to learn how to avoid it in the
future.

These statements (and some others in the thread) have been black and white
in placing all onus on the person accused of causing offense. The real
world is not black and white, and occasionally people take offense too
easily. I'll cite donglegate again. It's a worthwhile thought exercise to
consider, in light of the introduction of the CoC, how the community would
respond to an incident where almost everyone agrees that the perpetrator's
behavior is benign, and yet someone has taken serious offense to it. I
don't expect a concrete answer to this, it's just something to consider:
Just because someone is offended, doesn't *necessarily* mean that the
accused did something wrong, or that they're obliged to change their
behavior.

- Lang.

On Wed, Oct 14, 2015 at 2:12 PM, Chandler Carruth via llvm-dev <
llvm-dev at lists.llvm.org> wrote:

> On Wed, Oct 14, 2015 at 1:57 PM Philip Reames <listmail at philipreames.com>
> wrote:
>
>>
>>
>> On 10/14/2015 01:25 PM, Chandler Carruth via llvm-dev wrote:
>>
>> On Wed, Oct 14, 2015 at 1:02 PM Renato Golin <renato.golin at linaro.org>
>> wrote:
>>
>>> On 14 October 2015 at 20:35, Tanya Lattner via llvm-dev
>>> <llvm-dev at lists.llvm.org> wrote:
>>> > Related specifically to the developers meeting, we are growing quite
>>> > rapidly. For the past few years, we have been increasing our
>>> attendance from
>>> > 50 at the start to now over 350 attendees. With this many people (and
>>> many
>>> > new to the community), it seems important to have a code of conduct to
>>> refer
>>> > to and possibly *prevent* any incidents from happening. And for some
>>> people
>>> > (not sure exact percentage), it makes them feel more comfortable
>>> attending a
>>> > conference that has a code of conduct.
>>>
>>> Just an honest and simple question: would it make sense to have a
>>> different code of conduct for meetings and the rest?
>>>
>>> I know it sounds like a bad idea, but my rationale is that maybe this
>>> would at least solve some of the points that socially inept people
>>> feel pressure on the current proposal.
>>>
>>> Because the consequences of a physical meeting can be a lot tougher
>>> than any electronic one, and because timing is of the essence, the
>>> wording *has* to be stronger and an executive decision has to be
>>> implemented.
>>>
>>> But such strong wording and harsh unappealable consequences do make
>>> us, of the anti-social variety, very frightened. We grew in a world
>>> that never made sense, and we have suffered our childhoods and
>>> adulthoods in constant fear of irrational (to our minds) reprimands.
>>> This is not a simple matter, it's quite real and have made me
>>> seriously consider many times leaving the open source realm for good.
>>> I have left jobs and regressed in my career because of things like
>>> that.
>>>
>>> From the very wording in the proposed CoC, we don't want to leave
>>> anyone behind, including physical and mental disabilities. If that's
>>> true, and we really mean it, than imposing such a harsh CoC from the
>>> majority of opinions is exactly the opposite of that. People like me
>>> are clearly not the majority, the NAS UK estimates 1 every 100 people
>>> in England has some form of autism, but that's the whole point of a
>>> CoC, is to not forget about the people with some form of fragility.
>>>
>>
>> There is an extremely large difference between fragility and an inability
>> to be polite and respectful.
>>
>> Chandler, I think your opening here is a bit quick to dismiss Renato's
>> position.  Your following text is more reasonable, but you first sentence
>> comes across as a bit harsh.
>>
>
> I'm sorry it came across as harsh, and I'm not sure what led you to feel
> it was dismissive. I will try to be more clear here: I'm very literally
> trying to point out a significant difference between two concerns, neither
> of which can be dismissed, but which I think there are very different
> options for responding to.
>
>
>>
>> (For the record, I'd normally not have said anything, but since this is
>> specifically in a thread about community social norms...)
>>
>
> You should say something, even normally. =]
>
>
>>
>>
>> I do not think there is a useful way for us to encourage and welcome
>> individuals who, for whatever reason including medical reasons, are
>> literally *incapable* of interacting in a social setting in a civil,
>> polite, and respectful manner. That would be a no-win situation. But
>> reality is not this cut and dry or black and white.
>>
>> I have both friends and colleagues with autism and other severe mental,
>> social, and cultural challenges. And yet, they are not *incapable* of this.
>> Certainly, sometimes, it is a significantly greater challenge for them to
>> understand why people react in the way that they do. However, they take on
>> that challenge and learn and succeed at being wonderful people. Do they
>> have to work harder than I do? Some of them probably do. Do I try to
>> sympathize, remain patient, and help them as much as I can? Absolutely.
>> Does any of this mean it is *ok for them to be disrepectful?* Absolutely
>> not.
>>
>> I would read Renato's point as being in a round-about-way a request for
>> help.  How should he (or anyone) not familiar with the existing norms
>> within the community expect to function?
>>
>
> I mean, I did... and, much as you pointed out in another email, I think
> that a code of conduct is *exactly* the kind of tool that helps here? I'm
> not really certain what you're trying to say here.
>
>
>> Part of learning is making mistakes and being corrected.  Particular for
>> someone with a form of autism, those corrections may need to include an
>> explanation of what not to do again and why.  I think part of Renato's
>> concern - it definitely is part of mine! - is that he might say something,
>> unintentionally offend someone, and not get a chance to learn from it.
>>
>
> So, what gives you that impression from the proposed document? My reading
> is that there will be explicit feedback given, which would seem to directly
> help folks learn?
>
>
> There is a related issue that may be confusing matters. It is completely
> reasonable to desire and seek out information to help you understand what
> you can do to improve if communication goes poorly. I would hope that the
> advisory committee works very hard to provide this kind of feedback in
> every case where they can. However, it is incredibly important to not
> expect or demand that a person *you have made feel unsafe* take the time to
> explain why. Being forced to explain the problem can and in many cases does
> *exacerbate* the problem for the individual. If they are up for the
> challenge of explaining, wonderful. But if they are not, it is critical to
> respect that and give them the space to feel safe again. Talk to your
> friends, to others, to the advisory committee, to whomever else you need to
> to learn how to avoid it in the future.
>
> If why this is the case doesn't make sense, I can try to dig up resources
> that go into great detail on the psychological effects of being forced to
> help teach people why hostile behavior was hostile, but I don't have them
> handy at the moment. However, please trust me that this is a real and
> serious issue. As an example, in some cases it essentially forces people
> who are often put back into a victim mindset to relive whatever caused that
> mindset.
>
> Anyways, I don't *think* that is the issue here, but I wanted to provide
> the context in case it comes up in the future.
>
>
>> For the record, I have personally run into this in the past.  I've
>> managed to seriously offend a couple of folks and had *absolutely* no idea
>> why until a third party took me aside and explained what I did and how it
>> was perceived.  That doesn't change the fact that I'm still responsible for
>> having given offense or that I didn't do my best to make amends, but the
>> chance to learn without it being "game over" is key.
>>
>
> I don't see *anything* like a "game over" in the reporting guidelines. The
> decisions should be rational, measured, incremental, and something that can
> be questioned and re-examined as time goes on.
>
>
>>
>> Now, obviously, providing that learning opportunity should not be taken
>> too far.  If someone's safety is in question, "game over" is *absolutely*
>> the right response.  Nor does it mean that there can not be serious
>> consequences.  We simply need to keep in mind that behavior can change, and
>> that offense may not have been (probably wasn't) intentional.  If we keep
>> that in mind and steer towards moderation and informal correction (as we
>> have in the past), I don't see there being any inherent conflict here.
>>
>
> I think we're actually in total agreement here, and I think the document
> is too. If you see things that aren't, please point them out, and better
> yet suggest edits that would help.
>
> _______________________________________________
> LLVM Developers mailing list
> llvm-dev at lists.llvm.org
> http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20151014/cdcf174d/attachment.html>


More information about the llvm-dev mailing list