llvm dev - Oct 2015 - RFC: Introducing an LLVM Community Code of Conduct

Hi,

Given David's comments about an echo-chamber, I thought I might as well
throw my oar in and say I'm strongly in favour of this.

I'm not particularly fussed on the wordsmithing or the verbosity. In fact
one of the (unfortunate) primary uses of this document surely will be as
reference when a misbehavior occurs, so spelling at least some things out
in full might be useful in that case. The only concern I have on the
wordsmithing is if it became too much like the FreeBSD CoC - the example
thrown around has been "Don't make it personal, don't take it
personally"
which for me is way too terse and doesn't put enough onus on the aggresser
to be more polite in a cross-culture community. And that, in a nutshell, is
what separates our culture from that of some other projects.

I'm incredibly proud of the community to which I belong. PaulR mentioned
the number of women at our devmtgs as a (very rough) yardstick - that was
actually mentioned to me by someone who had attended a number of
conferences in the compiler/low-level-software field as being surprisingly
large in comparison to the others. I'm proud of that, and feel we need to
increase our inclusiveness.

I'm happy for this CoC to come in and codify the conventions of this
community. The way we treat people, especially newcomers, is exemplary in
my view and needs to continue. If this CoC can help that, so be it.

In fact, IMHO this thread has shown some of the hot-headedness that is
generally frowned upon in our community, so that make me even more in
favour of the CoC.

So +1 from me, and thanks for the hard work and the cat herding! :)

Cheers,

James



On Wed, 14 Oct 2015 at 07:29 Sean Silva via llvm-dev <
llvm-dev at lists.llvm.org> wrote:
> On Tue, Oct 13, 2015 at 12:44 PM, Tanya Lattner via llvm-dev <
> llvm-dev at lists.llvm.org> wrote:
>
>>
>> On Oct 13, 2015, at 11:16 AM, Renato Golin via llvm-dev <
>> llvm-dev at lists.llvm.org> wrote:
>>
>> On 13 October 2015 at 18:59, Chandler Carruth via llvm-dev
>> <llvm-dev at lists.llvm.org> wrote:
>>
>> We have *not* appointed any such committee at this point.
>> (...)
>> The appeal is to the board of the Foundation. I don't expect the
board to
>> *be* the committee here, quite the opposite.
>>
>>
>> This doesn't solve the problem. If the foundation appoints the
>> committee, appeals to the foundation are still open for abuse.
>>
>> Only democratic and transparent processes can work in this fashion,
>> and I don't see the foundation as being either.
>>
>>
>> I don't think that their job will be to impose moral authority, I
think
>> the
>> code of conduct is the basis they would be required to cite for any
>> decision. Their role should be much more focused on understanding what
has
>> happened, and ensuring it is responded to. I also think that is called
out
>> in the document.
>>
>>
>> So why the need to list the punishments and make sure that only
>> capital punishments can actually be appealed?
>>
>>
>> I'm surprised and saddened to hear you say this. I also don't
particularly
>> agree. I have interacted with almost every member of the foundation
board
>> as
>> a regular course of interacting with the community. The foundation is
>> completely handling the planning and running of the developer's
meeting.
>> Certainly, we're still in the infancy of figuring this stuff out,
but I
>> don't see a problematic lack of engagement.
>>
>>
>> Apart from the great work Tanya is doing with the LLVM meetings, I
>> don't absolutely anything coming from the foundation. Can you
>> elaborate?
>>
>> All in all, she was already doing great work before, so I take this
>> more as her personal merit than anything else.
>>
>> Since we're delving into this topic...
>>
>> I expected the foundation to own and improve the validation
>> infrastructure, and the web presence. So far I only saw Apple building
>> the great GreenBot, but when I wanted to help, it seemed this would be
>> something between me and on Apple employee.
>>
>> I also saw the migration to a new server (thanks again Tanya), but
>> that's very far from providing a stable infrastructure. If Tanya is
>> the only one doing things, of course she won't be able to do what
we
>> *need* in any reasonable time.
>>
>>
>> You are totally right. I am only one person. I do have help with
certain
>> aspects of the machine. What I’m aiming for is group of administrators
to
>> handle llvm.org. But I need to move all the infrastructure over to AWS
>> (I just moved the lists) and also separate out pieces of it (like perf)
to
>> another machine. I know that you have give me lots of advice in this
area,
>> and I’m sorry it hasn't been executed yet. There is a plan drafted
(with
>> input from more than just the board), but it has not been presented to
the
>> community yet.
>>
>> I expected the foundation sponsors to put in hardware and money to buy
>> servers, Amazon cloud instances, co-host testing devices with
>> universities in a way to foster inclusion and promote the project,
>> etc. Things that are obvious to me, but it seems got lost in the
>> foundation…
>>
>>
>> I’m sorry this is the perception. This is for sure the long term plan,
>> but many things have had to fall into place before . A lot of our
efforts
>> over the last year have been in actually establishing the Foundation as
a
>> whole. So this includes all the necessary legal, accounting, and other
>> pieces in place. These are all the things that are not at all
interesting
>> or really useful for the community, but they will be in the long term.
>>
>> I’d really like a better way to capture all of these great ideas. I’m
>> taking notes, but maybe we need another place (other than llvm-dev)
that
>> ideas can be shared.
>>
>>
>> I also don't see any transparency, not representation. We got a
memo
>> that it came into existence, than nothing else. I can't find any
>> document about the foundation except that blog post. As a non-profit,
>> I'd expect to be public memos, checks and balances, monthly
reports,
>> etc.
>>
>> This is entirely my fault (not the board as a whole) and its something
I
>> have been trying to rectify. You are exactly right. Everything we are
doing
>> should be 100% transparent. I have all of the board meeting minutes,
our
>> bylaws, everything that needs to go onto the website. Its something
that I
>> just didn’t make a high enough priority and I’m sorry about that. I
will
>> try to move this to the top of the list.
>>
>> If anyone wants copies of anything before then, I am more than happy to
>> send an email to you. Please do not hesitate to ask me.
>>
>
> Could you just throw stuff in the public LLVM SVN repo? Same way we have
> [lld] posts or [compiler-rt] posts on llvm-commits we would have
> [foundation] mailing list posts or whatever. Would also be nice in that
> llvm-commits is a natural avenue to disseminate information to involved
> developers.
>
> I'd definitely click into `[foundation] rXXXXXX - LLVM Foundation Oct
13
> 2015 meeting minutes` or whatever to take a look (that would be quite
> convenient for me actually).
>
> -- Sean Silva
>
>
>>
>> -Tanya
>>
>> I don't want to have those documents on my disk, I want them
public,
>> indexed by search engines, as soon as they're published.
>>
>> I can't see how the foundation represents my work in LLVM if they
>> essentially don't exist publicly, especially because there's no
way I
>> can influence. Representation is given, not taken.
>>
>>
>> cheers,
>> --renato
>> _______________________________________________
>> LLVM Developers mailing list
>> llvm-dev at lists.llvm.org
>> http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev
>>
>>
>>
>> _______________________________________________
>> LLVM Developers mailing list
>> llvm-dev at lists.llvm.org
>> http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev
>>
>> _______________________________________________
> 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/b251be50/attachment-0001.html>

Hi All,

 

I am also in favor of having a CoC. We may not particularly need it today, but
as the community grows and time goes by, we will unfortunately, but unavoidably,
require it --- and at that time, we will be happy to have it.

 

I would tend to prefer tersity over verbosity, as I think the spirit of the CoC
is more important than the actual enumerations, but that is a personal or
cultural preference, and I understand some people / culture may feel more
comfortable with a bit more verbosity.

 

I agree with all non-controversial polishing which have been discussed so far.

 

Thanks Chandler for herding the cats !

 

Cheers,

Arnaud

 

From: llvm-dev [mailto:llvm-dev-bounces at lists.llvm.org] On Behalf Of James
Molloy via llvm-dev
Sent: 14 October 2015 09:26
To: Sean Silva; Tanya Lattner
Cc: LLVM Dev; Chandler Carruth
Subject: Re: [llvm-dev] RFC: Introducing an LLVM Community Code of Conduct

 

Hi,

 

Given David's comments about an echo-chamber, I thought I might as well
throw my oar in and say I'm strongly in favour of this.

 

I'm not particularly fussed on the wordsmithing or the verbosity. In fact
one of the (unfortunate) primary uses of this document surely will be as
reference when a misbehavior occurs, so spelling at least some things out in
full might be useful in that case. The only concern I have on the wordsmithing
is if it became too much like the FreeBSD CoC - the example thrown around has
been "Don't make it personal, don't take it personally" which
for me is way too terse and doesn't put enough onus on the aggresser to be
more polite in a cross-culture community. And that, in a nutshell, is what
separates our culture from that of some other projects.

 

I'm incredibly proud of the community to which I belong. PaulR mentioned the
number of women at our devmtgs as a (very rough) yardstick - that was actually
mentioned to me by someone who had attended a number of conferences in the
compiler/low-level-software field as being surprisingly large in comparison to
the others. I'm proud of that, and feel we need to increase our
inclusiveness.

 

I'm happy for this CoC to come in and codify the conventions of this
community. The way we treat people, especially newcomers, is exemplary in my
view and needs to continue. If this CoC can help that, so be it.

 

In fact, IMHO this thread has shown some of the hot-headedness that is generally
frowned upon in our community, so that make me even more in favour of the CoC.

 

So +1 from me, and thanks for the hard work and the cat herding! :)

 

Cheers,

 

James

 

 

 

On Wed, 14 Oct 2015 at 07:29 Sean Silva via llvm-dev <llvm-dev at
lists.llvm.org> wrote:

On Tue, Oct 13, 2015 at 12:44 PM, Tanya Lattner via llvm-dev <llvm-dev at
lists.llvm.org> wrote:

 

On Oct 13, 2015, at 11:16 AM, Renato Golin via llvm-dev <llvm-dev at
lists.llvm.org> wrote:

 

On 13 October 2015 at 18:59, Chandler Carruth via llvm-dev
<llvm-dev at lists.llvm.org> wrote:



We have *not* appointed any such committee at this point.
(...)
The appeal is to the board of the Foundation. I don't expect the board to
*be* the committee here, quite the opposite.


This doesn't solve the problem. If the foundation appoints the
committee, appeals to the foundation are still open for abuse.

Only democratic and transparent processes can work in this fashion,
and I don't see the foundation as being either.





I don't think that their job will be to impose moral authority, I think the
code of conduct is the basis they would be required to cite for any
decision. Their role should be much more focused on understanding what has
happened, and ensuring it is responded to. I also think that is called out
in the document.


So why the need to list the punishments and make sure that only
capital punishments can actually be appealed?





I'm surprised and saddened to hear you say this. I also don't
particularly
agree. I have interacted with almost every member of the foundation board as
a regular course of interacting with the community. The foundation is
completely handling the planning and running of the developer's meeting.
Certainly, we're still in the infancy of figuring this stuff out, but I
don't see a problematic lack of engagement.


Apart from the great work Tanya is doing with the LLVM meetings, I
don't absolutely anything coming from the foundation. Can you
elaborate?

All in all, she was already doing great work before, so I take this
more as her personal merit than anything else.

Since we're delving into this topic...

I expected the foundation to own and improve the validation
infrastructure, and the web presence. So far I only saw Apple building
the great GreenBot, but when I wanted to help, it seemed this would be
something between me and on Apple employee.

I also saw the migration to a new server (thanks again Tanya), but
that's very far from providing a stable infrastructure. If Tanya is
the only one doing things, of course she won't be able to do what we
*need* in any reasonable time.

 

You are totally right. I am only one person. I do have help with certain aspects
of the machine. What I’m aiming for is group of administrators to handle
llvm.org. But I need to move all the infrastructure over to AWS (I just moved
the lists) and also separate out pieces of it (like perf) to another machine. I
know that you have give me lots of advice in this area, and I’m sorry it
hasn't been executed yet. There is a plan drafted (with input from more than
just the board), but it has not been presented to the community yet.





I expected the foundation sponsors to put in hardware and money to buy
servers, Amazon cloud instances, co-host testing devices with
universities in a way to foster inclusion and promote the project,
etc. Things that are obvious to me, but it seems got lost in the
foundation…

 

I’m sorry this is the perception. This is for sure the long term plan, but many
things have had to fall into place before . A lot of our efforts over the last
year have been in actually establishing the Foundation as a whole. So this
includes all the necessary legal, accounting, and other pieces in place. These
are all the things that are not at all interesting or really useful for the
community, but they will be in the long term.

 

I’d really like a better way to capture all of these great ideas. I’m taking
notes, but maybe we need another place (other than llvm-dev) that ideas can be
shared.

 


I also don't see any transparency, not representation. We got a memo
that it came into existence, than nothing else. I can't find any
document about the foundation except that blog post. As a non-profit,
I'd expect to be public memos, checks and balances, monthly reports,
etc.

This is entirely my fault (not the board as a whole) and its something I have
been trying to rectify. You are exactly right. Everything we are doing should be
100% transparent. I have all of the board meeting minutes, our bylaws,
everything that needs to go onto the website. Its something that I just didn’t
make a high enough priority and I’m sorry about that. I will try to move this to
the top of the list.

 

If anyone wants copies of anything before then, I am more than happy to send an
email to you. Please do not hesitate to ask me.

 

Could you just throw stuff in the public LLVM SVN repo? Same way we have [lld]
posts or [compiler-rt] posts on llvm-commits we would have [foundation] mailing
list posts or whatever. Would also be nice in that llvm-commits is a natural
avenue to disseminate information to involved developers.

 

I'd definitely click into `[foundation] rXXXXXX - LLVM Foundation Oct 13
2015 meeting minutes` or whatever to take a look (that would be quite convenient
for me actually).

 

-- Sean Silva

 

 

-Tanya





I don't want to have those documents on my disk, I want them public,
indexed by search engines, as soon as they're published.

I can't see how the foundation represents my work in LLVM if they
essentially don't exist publicly, especially because there's no way I
can influence. Representation is given, not taken.

 

cheers,
--renato
_______________________________________________
LLVM Developers mailing list
llvm-dev at lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev

 


_______________________________________________
LLVM Developers mailing list
llvm-dev at lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev

_______________________________________________
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/e2fe66a6/attachment.html>

On 10/14/2015 2:25 AM, James Molloy via llvm-dev wrote:
> The only concern I have on the wordsmithing is if it became too much
> like the FreeBSD CoC - the example thrown around has been "Don't
make it
> personal, don't take it personally" which for me is way too terse
and
> doesn't put enough onus on the aggresser to be more polite in a
> cross-culture community. And that, in a nutshell, is what separates our
> culture from that of some other projects.
The person who takes offense in something may do so under the influence 
of their own (potentially misguided) perceptions.  We cannot blindly put 
the focus on the "source" of the offending material and require them
to
conform to the idea of politeness shared by everyone.  This is how 
"political correctness" has come into existence.  The "don't
take it
personally" is excellent in the sense that it encourages a second look 
at the problem from a more impersonal distance.  This is something that 
would help in all forms of communication, not just in this particular case.

-Krzysztof

-- 
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, 
hosted by The Linux Foundation

On Wed, Oct 14, 2015 at 6:26 AM Krzysztof Parzyszek via llvm-dev <
llvm-dev at lists.llvm.org> wrote:
> On 10/14/2015 2:25 AM, James Molloy via llvm-dev wrote:
> > The only concern I have on the wordsmithing is if it became too much
> > like the FreeBSD CoC - the example thrown around has been
"Don't make it
> > personal, don't take it personally" which for me is way too
terse and
> > doesn't put enough onus on the aggresser to be more polite in a
> > cross-culture community. And that, in a nutshell, is what separates
our
> > culture from that of some other projects.
>
> The person who takes offense in something may do so under the influence
> of their own (potentially misguided) perceptions.  We cannot blindly put
> the focus on the "source" of the offending material and require
them to
> conform to the idea of politeness shared by everyone.  This is how
> "political correctness" has come into existence.  The
"don't take it
> personally" is excellent in the sense that it encourages a second look
> at the problem from a more impersonal distance.

I disagree.

I think it is entirely possible to find a way to communicate ideas about
compilers, optimizations, and software without offending anyone. In fact,
as is evidenced by our past track record here, in the overwhelming majority
of cases *we are already doing exactly this*.

So perhaps it is technically "unfair" to push people to be more
polite, but
all evidence is that it would require very little effort on their part than
has already been consistently put forth on the mailing lists. I think it is
very reasonable to ask for that on our mailing lists in order to make our
community more welcoming to those who have entrenched, specific reasons for
feeling ostracized by technical and open source communities in general. It
seems a small price to pay, and I pay it gladly.
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20151014/c10314d5/attachment.html>

Hi All,
> We should be polite and respectful, and insist upon that from others.
This is a wonderful summation.

.... the example thrown around has been "Don't make
it
>> personal, don't take it personally" which for me is way too
terse and
>> doesn't put enough onus on the aggresser to be more polite in a
>> cross-culture community....
>>
>
> The person who takes offense in something may do so under the influence of
> their own (potentially misguided) perceptions.  We cannot blindly put the
> focus on the "source" of the offending material and require them
to conform
> to the idea of politeness shared by everyone.  This is how "political
> correctness" has come into existence.  The "don't take it
personally" is
> excellent in the sense that it encourages a second look at the problem from
> a more impersonal distance.  This is something that would help in all forms
> of communication, not just in this particular case.

I agree that people may reactive sensitively, and occasionally overly
sensitively, to all sorts of fairly benign behavior (as in 'donglegate',
for example). Good judgement will always have to be employed by all
participants. I don't think that's incompatible with the introduction of
a
CoC, and the proposed CoC should not make people nervous about discussing
'sexting' on the dev-lists. The stated intent is to explicitly call out
behavior that we would consider inappropriate under our current (implicit)
community standards, and that seems like a worthwhile aim.

Cheers,
Lang.
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20151014/37f85a18/attachment.html>