Re: [gentoo-project] Why should you *not* vote on existing Council members

[desultory]

On 06/27/19 10:15, Rich Freeman wrote:
> Again, only speaking personally.  Also, on many of these issues we're
> just going to disagree on what the policy is and how it ought to apply
> - ultimately policy is up to Council.  Proctors just tries to apply
> the little guidance that exists in this area and accept whatever
> direction it is given.
>
> I'm just replying where something hasn't already been said..
>
> On Thu, Jun 27, 2019 at 1:23 AM desultory <desultory@g.o> wrote:
>>
>> On 06/26/19 08:36, Rich Freeman wrote:
>>> On Wed, Jun 26, 2019 at 12:24 AM desultory <desultory@g.o> wrote:
>>>>
>>>> On 06/24/19 06:55, Rich Freeman wrote:
>>>>> Speaking only for my personal opinion:
>>>>>
>>>> by
>>>> your espoused reasoning even filing a bug requesting ComRel intervention
>>>> due to demonstrable violation of the CoC would not be a suitable venue
>>>> for arguably negative comments regarding an individual, as (per ComRel
>>>> [ComRel]) there is no indication that such bugs would necessarily be
>>>> private.
>>>
>>> There is also no indication that such bugs would be public.  IMO the
>>> ComRel policy should define expectations of privacy because this has
>>> been a problem in the past with Council appeals, which I believe I
>>> have commented on the lists about previously.
>>>
>> So no ComRel needs to clean house in order to avoid running afoul of a
>> new (unilaterally conceived) policy that runs distinctly counter to
>> existing practice.
>
> I was saying that ComRel needed to document expectations of privacy
> back when I was on Council, which was a while ago, and long before
> Proctors was restarted.
>
> People share all kinds of sensitive stuff with Comrel.  It absolutely
> should be clear whether they can expect it to remain private.  Having
> heard Comrel appeals I can tell you what is wrong with the current
> state.
>
> A few witnesses share sensitive concerns with Comrel about some dev
> with the understanding that it will be kept private (likely given by
> private assurance). Comrel ends up taking action against the dev.  The
> dev appeals to Council.  The Council gets a dump of all the evidence,
> much of which is sensitive, and where promises about privacy have
> already been made.  What does Council do?
>
> Council could decide that we don't act on private info, but then what
> was the point in soliciting it in the first place?  Also, Council has
> good reason to think that somebody bad is going on, which they are now
> ignoring.
>
> Council could decide to uphold the Comrel action, and keep the private
> info private.  Now you get the usual conspiracy theories about secret
> cabals running Gentoo, and no official policies one way or another.
>
> Council could uphold the Comrel action, and then publish the private
> info.  Now you get people really upset about broken promises on
> sensitive issues.
>
> No matter what the policy is set to, somebody will be upset.  However,
> we should still have a published policy.
>
Saying something as a member of the council does not mean implementing
something as a member of the council, though only little effort
separates them. Blaming ComRel for what others did not do and for not
setting policy for the council, from which it draws its mandate and to
which it reports (not dictates), are both absurd.

> This is why Proctors has a published policy.  Proctors deals with
> stuff done in public on Gentoo communications media, on lists, bugs,
> IRC, etc.  Proctors doesn't deal with stuff that happens in private,
> anywhere.  Proctors doesn't accept private evidence - anything
> submitted will go in a public bug.  Anybody with a concern about what
> Proctors is doing can go search on bugzilla and see the same things we
> see.  The most ephemeral stuff we deal with would be unlogged but
> official IRC channels, but usually plenty of people have personal logs
> they can look at for these, and even then we rarely get involved
> because most are already moderated.
>
Having "a published policy" tends to imply something more than literally
[proctors] :
"Note: All proctors matters will be tracked in public bugs, including
all communications sent in the request. The scope of proctors actions is
limited to activities on public communications media, so there is no
expectation of privacy around the handling of these issues."
embedded in a document, it seems entirely fair to describe that as more
akin to "the fine print" than "a published policy" as there is indeed no
reference to having everything related to all proctors decisions kept
public in the policy description itself. Rather the *contrary* in fact
given that the note claims that *all* proctors matters will be tracked
in public bugs while an "important" highlighted section of the section
describing disciplinary actions indicates that the proctors will not
necessarily have any record at all of instances where they were asked to
intervene:
"Before applying any of the following disciplinary policies, the
Proctors team will try to discuss the problem with the offender in order
to solve it in a more peaceful way. However, it is possible for the
Proctors to apply the penalty without further discussions in severe CoC
violations (direct attacks, insults, name-calling etc)."
Thus by the implication of proctors own published policy the publication
of all proctors data is less important than contacting the subject of a
complaint to resolve the matter before disciplinary action is taken,
curiously there is no indication that this was carried through in the
instance which spawned this discussion.

> Bugzilla in general is within scope of Proctors.  We would generally
> not deal with Comrel bugs because:
> 1. Comrel already has all the powers Proctors has to deal with CoC
> issues.  We don't add value.
If proctors do not add value in areas in which ComRel operates and
ComRel operates everywhere that proctors do, how is the existence of
proctors justified? Is a theoretically neutral party valueless in regard
to helping keep tempers in check in the very circumstance in which one
would expect them to be most strained? Is maintaining civil discourse
outside of the purview of proctors?

> 2. Most of these bugs are generally hidden, and stuff that isn't
> public generally isn't our scope.
Given that you have repeatedly noted that proctors sole area of
responsibility is public media, something which is only mentioned in the
note I quoted above, how would a hidden bug fall within proctors scope
at all?

> 3. Comrel IS an appropriate forum for frank discussion of a lot of
> stuff that would violate the CoC in public, so it requires a different
> approach, which is what Comrel already specializes in.
>
Given the list of unacceptable behaviors enumerated in the CoC [CoC],
specifically:
 * Flaming and trolling.
 * Posting/participating only to incite drama or negativity rather than
to tactfully share information.
 * Being judgmental, mean-spirited or insulting.
 * Constantly purveying misinformation despite repeated warnings.

What productive value, exactly, would any of that add to a discussion of
disciplinary action or policy?

>> Which would make this yet another new policy spawned for no
>> evident reason other than having not thought through the implications of
>> an existing policy, and not bothering to consult with affected parties,
>> again.
>
> I don't hear Comrel complaining.  I'd be shocked if they had concerns
> over what we're doing, and have a liason on our team for just that
> reason.  It isn't like we make stuff up in a vacuum.
>
I find it distinctly curious that you infer that ComRel would
necessarily be the aggrieved party in regards to comments made in ComRel
bugs, as opposed to individuals subject to them, filing them, or
otherwise involved in them. Indeed, ComRel not having any concerns here
could itself raise concerns about how ComRel.

As for the statement that proctors do not "make stuff up in a vacuum",
that seems dubious given the freshly espoused "no arguably negative
comments regarding individuals" policy which had no advance notice prior
to it being enforced.

>>> While it isn't Gentoo's policy I'd suggest taking a look at the FSF's
>>> CoC.  It does a decent job (IMO) of explaining why personal attacks
>>> are counterproductive even if you think they should be allowed, which
>>> they are not (at least not unless Council says otherwise).
>>>
>> The most obvious problem with that is what was warned about was not a
>> personal attack. There is a difference between "$person lack empathy"
>> and "$person is an inhuman monster" which seems to have been lost in
>> this action by proctors.
>
> So, I think we're arguing over the definition of "personal attack" -
> either statement is not appropriate on our lists.
>
One is constructive feedback, the other is not. Banning constructive
feedback is definitionally not constructive. To treat such a ban as
being in effect prior to announcing such a ban is absurd.

> And we certainly do make distinctions, which is why only a warning was issued.
>
A warning for violating a policy which did not exist prior to that warning.

>> This is yet another occurrence of a fallacy which is distressingly
>> common in various media: since one does not think that a system is being
>> overly abused now, why should anyone be at all concerned about abuses in
>> the system? Especially when that system is, in the instance in question,
>> being abused.
>
> If Council feels that our action was inappropriate they can take
> whatever action they feel is necessary.  We enforce the CoC as we
> believe it was intended to be enforced.  A few people have voiced
> disagreements, which is to be expected.  That is why we elect Council
> members.
>
Duly noted. Apparently, I will need to inquire with the council.

> All Proctors actions and bugs (whether action is taken or not) are
> public.  Anybody can review what we're doing and raise whatever
> concerns they wish, as you have done.
>
By the proctors own published policy, cited above, that claim is false.
Yes, the bugs are, by policy public, but all actions taken and not
cannot possibly be documented, please do not overgeneralize.

> IMO it is a good system.  There will be disagreements, but I think
> this is about as transparent a system as I can think of, and
> suggestions for improvement are always welcome.
>
Consistency in enforcement actions and publishing policies before
enforcing them would be a start, but considering that the proctors
project has been running for roughly a year at this point "a start"
seems rather late.

>> It is entirely possible to make policies which are unclear, or
>> inexplicit on some point, which are then taken, incorrectly, to imply
>> something which is itself then made policy and enforced.
>
> Of course.  And that is why we have the opportunity for feedback.  All
> policies need refinement over time, and Council is the appropriate
> place to bring concerns about the meaning of the CoC.
>
And enforcing bodies are, or at least should be, suitable points of
contact for concerns regarding their handling of the CoC.

>> There have been
>> numerous personal attacks on the lists in the time since the proctors
>> project was started (restarted, if you prefer), none received a warning
>> for months, then a critique of how someone handles one of their roles
>> was treated as an actionable violation.
>
> True.  I do not claim that this was the worst violation since Proctors existed.
>
Yet it was the only enforcement action. Why?

> In general we avoid opening bugs every time a minor issue comes up.
> However, we didn't open this bug.  Once a complaint was submitted to
> us our options were basically to close it without action, or close it
> taking some kind of action.  You might disagree with the decision we
> made in this case, but IMO it was the right one.
>
Your argument appears to be essentially that you were bound to act
because a bug was filed, while proctors policy explicitly states that it
has the option to not enforce it own policies even when it considers a
violation to have occurred:
"The following disciplinary actions may or may not be enforced when the
Proctors become aware of a direct CoC violation."
Please explain.

> I'm not suggesting that I want everybody to go opening up Proctors
> bugs everytime somebody does something you don't like.  However, if
> bugs are opened, we're going to follow our process to resolve them,
> and we will generally do so quickly.  Over time I'm sure we'll both
> get better at it, and people will become more used to how these are
> being handled, and maybe we'll have fewer CoC violations in the first
> place.
>
Given that you left out my question which you appear to be tangentially
addressing there, allow me to further clarify: I was not addressing
whether or not I "liked" your question, I was addressing whether or not
it complied with the CoC. By your own stated standards it would appear
to not comply. Thus my question stands: is filing a proctors bug about
such questions an intended effect of your newly espoused policy? Would
it be handled as the newly espoused policy and recent warning would
indicate?

Further, would having a warning issued by proctors constitute failure to
comply with the prerequisites for proctors membership? Specifically:
"A Proctor must be a Gentoo developer for at least 1 year and during
this time must have demonstrated good behavior."

>> I think that it is distinctly unrealistic to treat a personal critique
>> as an actionable personal attack when personal attacks are regularly
>> ignored by proctors.
>
> So, you can't get out of a speeding ticket by arguing that the police
> failed to pull over EVERY car that was speeding.  No CoC enforcement
> will be perfect, nor is it intended to be really.  The goal is to
> steer things in the right direction, and nudges over time will
> hopefully get things going in the right direction.  This is why I am
> emphatic that warnings should not be seen as reflecting on the
> individual.  Getting a warning doesn't mean that you're the worst
> person in Gentoo - it just means that you did something wrong, and you
> should try not to do it again.  That's it, and if we actually heed the
> warnings maybe it will be a nicer community to participate in.
>
Conversely, the police cannot argue that they are enforcing speed limits
effectively if they ticket only one car per year, and when asked about
why they only ticketed that one car instead of the dozens of others that
went past, uncited, at much higher speeds proclaim that they could not
catch the cars they photographed and logged as having more flagrantly
violated the speed limit but they could catch this car that was just
barely exceeding the speed limit, thus speed limits are enforced
effectively. The populace in general would also, rather likely and quite
rightly, be rather taken aback if the police upon ticketing this one
driver announced that speed limits would thenceforth be enforced such
that if any part of a vehicle exceeds the posted limit, the driver would
be fined for traveling at the speed of the fastest part of the vehicle.

You keep arguing that the proctors project has only been around for a
year and that it is somehow just starting out, the very idea of the
proctors rather strongly implies the opposite: members of the team
should have some sense of what they are to be doing before they ever
become proctors. Furthermore, proctors should not act in a capricious
manner with regard to their duties, to do otherwise is to destroy, or at
very least debase, the value of the role.

[proctors] https://wiki.gentoo.org/wiki/Project:Proctors
[CoC] https://wiki.gentoo.org/wiki/Project:Council/Code_of_conduct