code-of-conduct: Remove explicit list of discrimination factors
diff mbox series

Message ID 20181017071902.30102-1-geert@linux-m68k.org
State New, archived
Headers show
Series
  • code-of-conduct: Remove explicit list of discrimination factors
Related show

Commit Message

Geert Uytterhoeven Oct. 17, 2018, 7:19 a.m. UTC
Providing an explicit list of discrimination factors may give the false
impression that discrimination based on other unlisted factors would be
allowed.

Avoid any ambiguity by removing the list, to ensure "a harassment-free
experience for everyone", period.

Fixes: 8a104f8b5867c682 ("Code of Conduct: Let's revamp it.")
Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>
---
The use of "race" may also conflict with the United Nation's views on
this matter, cfr. e.g. the UNESCO's "Four statements on the race
question"[1][2] and "Declaration on Race and Racial Prejudice"[3].

[1] https://en.wikipedia.org/wiki/The_Race_Question
[2] http://unesdoc.unesco.org/images/0012/001229/122962eo.pdf
[3] http://portal.unesco.org/en/ev.php-URL_ID=13161&URL_DO=DO_TOPIC&URL_SECTION=201.html
---
 Documentation/process/code-of-conduct.rst | 5 +----
 1 file changed, 1 insertion(+), 4 deletions(-)

Comments

Josh Triplett Oct. 17, 2018, 9:13 a.m. UTC | #1
On Wed, Oct 17, 2018 at 09:19:01AM +0200, Geert Uytterhoeven wrote:
> Providing an explicit list of discrimination factors may give the false
> impression that discrimination based on other unlisted factors would be
> allowed.

This impression is, in fact, false, as has already been discussed
elsewhere. I had hoped that discussion would suffice.

As mentioned there: The original commit explicitly said "Explicit
guidelines have demonstrated success in other projects and other areas
of the kernel."; this is precisely the kind of explicit guideline it
refers to. Listing explicit cases to cover does not imply other cases
are not covered; it does, however, ensure that the listed cases *are*,
and helps people know that they're covered.

This patch is not OK, and defeats one of the purposes of the original
change.
Geert Uytterhoeven Oct. 17, 2018, 9:31 a.m. UTC | #2
Hi Josh,

Thanks for your comments!

On Wed, Oct 17, 2018 at 11:13 AM Josh Triplett <josh@joshtriplett.org> wrote:
> On Wed, Oct 17, 2018 at 09:19:01AM +0200, Geert Uytterhoeven wrote:
> > Providing an explicit list of discrimination factors may give the false
> > impression that discrimination based on other unlisted factors would be
> > allowed.
>
> This impression is, in fact, false, as has already been discussed
> elsewhere. I had hoped that discussion would suffice.

The CoC FAQ is not part of the CoC, and not part of the Linux kernel.
If the CoC is imprecise, it should be fixed in the CoC, not in a separate
document hosted elsewhere, as discussed elsewhere.

Comparison with the GPL and the GPL FAQ is not appropriate, as the GPL
is still the precise legal document, while its FAQ is a clarification using
laymen's terms.

> As mentioned there: The original commit explicitly said "Explicit
> guidelines have demonstrated success in other projects and other areas
> of the kernel."; this is precisely the kind of explicit guideline it

Given the original commit was not submitted for and objected to public
review, nobody had the chance to question these statements, and ask for
pointers of proof, which would surely have happened.

> refers to. Listing explicit cases to cover does not imply other cases
> are not covered;

It does, if not accompanied by "examples of...", like in the other sections.

> it does, however, ensure that the listed cases *are*,
> and helps people know that they're covered.

So you agree people cannot know if the unlisted cases are covered or not?

> This patch is not OK, and defeats one of the purposes of the original
> change.

So the purpose of the original change was to list a number of factors,
without saying that it was just a list of examples?

Gr{oetje,eeting}s,

                        Geert
Guenter Roeck Oct. 17, 2018, 1:32 p.m. UTC | #3
On 10/17/2018 02:31 AM, Geert Uytterhoeven wrote:
> Hi Josh,
> 
> Thanks for your comments!
> 
> On Wed, Oct 17, 2018 at 11:13 AM Josh Triplett <josh@joshtriplett.org> wrote:
>> On Wed, Oct 17, 2018 at 09:19:01AM +0200, Geert Uytterhoeven wrote:
>>> Providing an explicit list of discrimination factors may give the false
>>> impression that discrimination based on other unlisted factors would be
>>> allowed.
>>
>> This impression is, in fact, false, as has already been discussed
>> elsewhere. I had hoped that discussion would suffice.
> 
> The CoC FAQ is not part of the CoC, and not part of the Linux kernel.
> If the CoC is imprecise, it should be fixed in the CoC, not in a separate
> document hosted elsewhere, as discussed elsewhere.
> 
> Comparison with the GPL and the GPL FAQ is not appropriate, as the GPL
> is still the precise legal document, while its FAQ is a clarification using
> laymen's terms.
> 
>> As mentioned there: The original commit explicitly said "Explicit
>> guidelines have demonstrated success in other projects and other areas
>> of the kernel."; this is precisely the kind of explicit guideline it
> 
> Given the original commit was not submitted for and objected to public
> review, nobody had the chance to question these statements, and ask for
> pointers of proof, which would surely have happened.
> 
>> refers to. Listing explicit cases to cover does not imply other cases
>> are not covered;
> 
> It does, if not accompanied by "examples of...", like in the other sections.
> 
>> it does, however, ensure that the listed cases *are*,
>> and helps people know that they're covered.
> 
> So you agree people cannot know if the unlisted cases are covered or not?
> 
>> This patch is not OK, and defeats one of the purposes of the original
>> change.
> 
> So the purpose of the original change was to list a number of factors,
> without saying that it was just a list of examples?

One could consider adding something like "discrimination factors such as",
or maybe "or any other discrimination factors not listed here" to the
original text. Or a simple "regardless of, for example, ...".

Guenter
Guenter Roeck Oct. 17, 2018, 1:45 p.m. UTC | #4
On 10/17/2018 02:13 AM, Josh Triplett wrote:
> On Wed, Oct 17, 2018 at 09:19:01AM +0200, Geert Uytterhoeven wrote:
>> Providing an explicit list of discrimination factors may give the false
>> impression that discrimination based on other unlisted factors would be
>> allowed.
> 
> This impression is, in fact, false, as has already been discussed
> elsewhere. I had hoped that discussion would suffice.
> 
> As mentioned there: The original commit explicitly said "Explicit
> guidelines have demonstrated success in other projects and other areas
> of the kernel."; this is precisely the kind of explicit guideline it
> refers to. Listing explicit cases to cover does not imply other cases
> are not covered; it does, however, ensure that the listed cases *are*,
> and helps people know that they're covered.
> 

That is really a matter of opinion. Mathematically speaking, your statement
is incorrect. One may wonder why the list is made explicit without hint
that it is an example. For example, political or social views are _not_
listed. Wasn't the same CoC used in other projects to at least try to
punish individuals with specific political and/or social opinions,
just for having those opinions and expressing them outside the scope of
the project ?

> This patch is not OK, and defeats one of the purposes of the original
> change.

The CoC, as it stands, singles out maintainers for enforcement action.
Based on your statement, is it correct to assume that this was on
purpose ? If not, what is the explicit list of purposes of the
original change ?

Guenter
Josh Triplett Oct. 17, 2018, 3:21 p.m. UTC | #5
On Wed, Oct 17, 2018 at 11:31:35AM +0200, Geert Uytterhoeven wrote:
> Hi Josh,
> 
> Thanks for your comments!
> 
> On Wed, Oct 17, 2018 at 11:13 AM Josh Triplett <josh@joshtriplett.org> wrote:
> > On Wed, Oct 17, 2018 at 09:19:01AM +0200, Geert Uytterhoeven wrote:
> > > Providing an explicit list of discrimination factors may give the false
> > > impression that discrimination based on other unlisted factors would be
> > > allowed.
> >
> > This impression is, in fact, false, as has already been discussed
> > elsewhere. I had hoped that discussion would suffice.
> 
> The CoC FAQ is not part of the CoC, and not part of the Linux kernel.

I wasn't referring just to that; I'm referring to the discussion we've
already had on this exact point.

> > refers to. Listing explicit cases to cover does not imply other cases
> > are not covered;
> 
> It does, if not accompanied by "examples of...", like in the other sections.

"for everyone, regardless of ..." still says "for everyone", making the
"regardless of ..." inherently a non-exhaustive list of factors.

> > it does, however, ensure that the listed cases *are*,
> > and helps people know that they're covered.
> 
> So you agree people cannot know if the unlisted cases are covered or not?

People in underrepresented and commonly marginalized groups, especially
those more commonly overlooked, don't always know if a given group has
taken their particular group into account or given any thought to it.
Explicit inclusion helps, and this is a standard guideline often cited
for good codes of conduct.

That doesn't make other groups *not* covered. But *if* there is a
particular commonly marginalized group that you feel this should
*explicitly* cover and doesn't, I'd suggest *adding* that group rather
than deleting the existing effort to be explicitly inclusive. (And
again, I'd suggest doing so upstream first.)

> > This patch is not OK, and defeats one of the purposes of the original
> > change.
> 
> So the purpose of the original change was to list a number of factors,
> without saying that it was just a list of examples?

You seem to be actively trying to read something more into what I said.
One of the key purposes of the original change was to make the kernel a
"a welcoming environment to participate in", and to provide "explicit
guidelines".
Josh Triplett Oct. 17, 2018, 3:22 p.m. UTC | #6
On Wed, Oct 17, 2018 at 06:32:36AM -0700, Guenter Roeck wrote:
> One could consider adding something like "discrimination factors such as",
> or maybe "or any other discrimination factors not listed here" to the
> original text. Or a simple "regardless of, for example, ...".

These sound like perfectly reasonable ways to address this. Please
consider submitting a patch upstream based on this reasoning.
James Bottomley Oct. 17, 2018, 3:49 p.m. UTC | #7
On Wed, 2018-10-17 at 08:21 -0700, Josh Triplett wrote:
> People in underrepresented and commonly marginalized groups,
> especially those more commonly overlooked, don't always know if a
> given group has taken their particular group into account or given
> any thought to it. Explicit inclusion helps, and this is a standard
> guideline often cited for good codes of conduct.

Actually, that's not a good thing to do in a vacuum: you have to be
really careful about how you do this from a legal point of view.  The
argument over whether enumerating specific rights or classes disparages
others has been going on for centuries.  To give you an example of how
far back it goes: it's the reason for the ninth amendment to the US
constitution.

The commonly accepted legal way of doing this today is the statement

"examples of X include but are not limited to: ..."

which is thought to work in most jurisdictions and is what you'll find
in all US corporate codes of conduct or ethics.

James
Josh Triplett Oct. 17, 2018, 4 p.m. UTC | #8
On Wed, Oct 17, 2018 at 08:49:15AM -0700, James Bottomley wrote:
> On Wed, 2018-10-17 at 08:21 -0700, Josh Triplett wrote:
> > People in underrepresented and commonly marginalized groups,
> > especially those more commonly overlooked, don't always know if a
> > given group has taken their particular group into account or given
> > any thought to it. Explicit inclusion helps, and this is a standard
> > guideline often cited for good codes of conduct.
> 
> Actually, that's not a good thing to do in a vacuum: you have to be
> really careful about how you do this from a legal point of view.  The
> argument over whether enumerating specific rights or classes disparages
> others has been going on for centuries.  To give you an example of how
> far back it goes: it's the reason for the ninth amendment to the US
> constitution.
> 
> The commonly accepted legal way of doing this today is the statement
> 
> "examples of X include but are not limited to: ..."
> 
> which is thought to work in most jurisdictions and is what you'll find
> in all US corporate codes of conduct or ethics.

Which is a much better proposal than removing the list entirely.
Joe Perches Oct. 17, 2018, 4:18 p.m. UTC | #9
On Wed, 2018-10-17 at 02:13 -0700, Josh Triplett wrote:
> On Wed, Oct 17, 2018 at 09:19:01AM +0200, Geert Uytterhoeven wrote:
> > Providing an explicit list of discrimination factors may give the false
> > impression that discrimination based on other unlisted factors would be
> > allowed.
> 
> This impression is, in fact, false, as has already been discussed
> elsewhere.

The use of fact above is instead merely an assertion.

>  I had hoped that discussion would suffice.

Nope.

> As mentioned there: The original commit explicitly said "Explicit
> guidelines have demonstrated success in other projects and other areas
> of the kernel."

Correlation etc...
Mark Brown Oct. 17, 2018, 6:36 p.m. UTC | #10
On Wed, Oct 17, 2018 at 08:21:02AM -0700, Josh Triplett wrote:
> On Wed, Oct 17, 2018 at 11:31:35AM +0200, Geert Uytterhoeven wrote:
> > On Wed, Oct 17, 2018 at 11:13 AM Josh Triplett <josh@joshtriplett.org> wrote:

> > > it does, however, ensure that the listed cases *are*,
> > > and helps people know that they're covered.

> > So you agree people cannot know if the unlisted cases are covered or not?

> People in underrepresented and commonly marginalized groups, especially
> those more commonly overlooked, don't always know if a given group has
> taken their particular group into account or given any thought to it.
> Explicit inclusion helps, and this is a standard guideline often cited
> for good codes of conduct.

I have heard some complaints that the strong push to include these lists
has ended up devaluing them, it becomes hard for people to tell if the
list is just a cut'n'paste job or if the people responsible for the code
of conduct really understand the issues affecting the groups they
include and it can be extra disappointing if there are problems.  I
particularly remember a friend of mine getting into an argument with a
conference being hosted somewhere where being gay was a capital offence
questioning the inclusion of sexuality on their list, it seemed fairly
clear that the organizers meant well and were trying to do the right
thing but weren't really aware.

This doesn't mean don't try but it's definitely a factor to consider,
especially when using an off the shelf code of conduct - there's just
never going to be a single right answer with a lot of this stuff.
Pavel Machek Oct. 22, 2018, 9:06 p.m. UTC | #11
On Wed 2018-10-17 09:19:01, Geert Uytterhoeven wrote:
> Providing an explicit list of discrimination factors may give the false
> impression that discrimination based on other unlisted factors would be
> allowed.
> 
> Avoid any ambiguity by removing the list, to ensure "a harassment-free
> experience for everyone", period.
> 
> Fixes: 8a104f8b5867c682 ("Code of Conduct: Let's revamp it.")
> Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>

Acked-by: Pavel Machek <pavel@ucw.cz>

[I'd preffer going to old code of conflict. But even this is an improvement.]

									Pavel

Patch
diff mbox series

diff --git a/Documentation/process/code-of-conduct.rst b/Documentation/process/code-of-conduct.rst
index ab7c24b5478c6b30..e472c9f86ff00b34 100644
--- a/Documentation/process/code-of-conduct.rst
+++ b/Documentation/process/code-of-conduct.rst
@@ -6,10 +6,7 @@  Our Pledge
 
 In the interest of fostering an open and welcoming environment, we as
 contributors and maintainers pledge to making participation in our project and
-our community a harassment-free experience for everyone, regardless of age, body
-size, disability, ethnicity, sex characteristics, gender identity and
-expression, level of experience, education, socio-economic status, nationality,
-personal appearance, race, religion, or sexual identity and orientation.
+our community a harassment-free experience for everyone.
 
 Our Standards
 =============